« first day (4811 days earlier)      last day (151 days later) » 

1:08 AM
We've spent a lot of effort on investigations into supporting new askers but we've done little to help people writing answers. In the grand scheme of problematic answers, though, I'd rank addressing comment answers ("I have the same issue, are there any updates?!?!?) would be much more impactful. That's not to say this isn't worthy, just that it's one of many issues and piecemeal solutions aren't the direction we want to go. — Catija ♦ 6 mins ago
@Catija agreed, though nit: "I have the same issue, are there any updates" is debatable as a comment- I find it hard to classify differently than a "+1", which is explicitly discouraged in "when not to comment" in /help/privileges/comment. The uncontroversial better option would be to upvote. By the time you get comment privs, you can upvote already.
 
@starball bah. You know what I mean.
People treating SO like a forum is the issue.
 
@Catija hehe
 
I managed to solve an issue with my browser on SU and as part of my research for the question I found a bug report on the Google and most of the comments were people who had the issue and were trying to figure out why it was marked as fixed.
 
I've kind of entertained the idea in my mind of streaming my curation activities online. but it would be my first time streaming anything, and I kind of am not sure about the tradeoff between educating people and the effort of explaining why I'm doing what I'm doing (and the repetitiveness of it) and possible backlash (Ex. you jerkface!!! this person just wants to know if a solution has been found!!)
@Catija I've seen similar happen on VS Code's issue tracker / issue trackers of VS Code extensions. maintainers think they fixed a bug / they did fix the bug, but either it didn't get fixed completely, or people have a new similar issue.
there are also less helpful oopsies I've seen maintainers do. A funny recent one: maintainer closes ticket as a duplicate of itself.
Stack Overflow's anti-"+1"/"bump"-ness is pretty great. bottom section of stackoverflow.com/a/77232931/11107541 gives a pretty good example of what things can happen in the extremes on issue trackers that don't have cleaning/denoising mechanisms like there are on SO.
 
@starball I haven't ever seen someone streaming that sort of thing but the people I've seen on YouTube making Minecraft videos seem to frame the Vo as if they're streaming but it's pretty clear as an adult watching, that they have a script the prepare after editing.
 
1:19 AM
@Catija what's a "Vo"? And is this a suggestion to script a voiceover over a recording? because I haven't thought of that option. I'll mull over it.
also, what minecraft youtubers do you watch? I've been watching hermitcraft people for the past year-ish
 
@starball I can see some value in it, as someone who is trying to figure out the cause of a bug. The bug itself in this case was a couple of years old but seeing periodic indication that people are still experiencing the problem made me feel more confident about it still being a general issue. Who knows whether my solution (third-party cookies) is the answer they're looking for.
@starball VO. I missed out on the capitalization. Voice over, so you talking while working.
@starball my kids mostly watch them. I don't remember them all but the one we go back to consistently is ACookieGod.
 
@Catija ah I didn't think of that. but what I'd suggest people to do instead of just "I have this too" is to actually post their version and environment/setup info. Then you have some concrete info about conditions for reproduction.
@Catija ah
 
The benefits of a recording rather than streaming is you can remove the mundane / less interesting cases that might lead people to wander off. I'm guessing a lot of what you do is pretty rote or repetitive. I also watch DGR Dave... I've seen one stream but aside from the opportunity for live interaction, it felt really slow. The recordings are more engaging.
 
@Catija good tip.
 
That said, there's the question of whether people will be interested in "how I curate SO"... I think a series of videos about specific curation tasks and best practices where you focus on one aspect (say suggested edit reviews) and cover the general case and some variations that come up along the way. Recording the process and rearranging the recording into a story arc can be really helpful.
I've occasionally made recordings of bugs I can reproduce or reviews of new UI workflows so that people can see what's going on or where I have thoughts. It's helpful to talk through things while recording so you can illustrate the steps clearly. That'd be really important for you so you have time in the video to explain in the VO.
I also really love my cursor decoration because it is much easier for people to follow what I'm pointing at.
 
1:41 AM
The more I think about making videos teaching about SE usage, the more I have doubts about whether it's worth it :P
topic-wise, I think tips on using search and just plain-reading of some Help Center pages on asking could be valuable (for people who don't like reading)
part of the hesitation for me is just not being sure what value I could add in video form over what can be conveyed in text. the whole don't-use-images-of-text thing has kind been pretty deeply ingrained into me by this point.
 
2:01 AM
@starball the value is that a lot of people now really connect with video and absorb content more easily than through reading. I don't understand it myself... give me a bullet list and some screenshots and I'm way better off but so many guides now are switching to video.
 
 
4 hours later…
6:15 AM
7 messages moved to Chimney
 
 
2 hours later…
7:45 AM
@starball You would have to very careful not to expose any PII when streaming ...
 
8:27 AM
@DavidPostill you mean of myself right?
I have a browser profile that I can use just for SE things, so I'd be using that. and I'm pretty sure most screen-recording software on web and desktop let you scope to a specific window or virtual desktop.
 
9:03 AM
@Catija on the other side, the fact that older users try to make the interactions between users more and more ... aseptic contributes to destroy any felling of community imho.
Just wondering, do you have any data about the chat usage?
I think it greatly dropped in the last few years, right?
 
9:42 AM
well, back to my rants against Crunchyroll I go.
 
 
2 hours later…
11:31 AM
Isn't the whole network an adult section of the internet?
 
That is true.
 
12:07 PM
@SPArcheon I think it depends on the site, which is why I said SO specifically. I don't see the volume of problematic "answers" on other sites or the formulation of what's wrong with the "answer" is different. I do feel like the responses to these new users can feel very much like white blood cells seeking out invaders to eradicate... but I understand why, so it's a complicated situation.
When you see the same thing over and over, even knowing that it's different people you're talking to, your patience gets shorter and shorter over time and you don't necessarily even realize how you're coming across... and a lot of that frustration is on us for not preventing the problem through better onboarding. So while I agree, I think the root cause falls back to us to address.
 
@starball No, I meant any PII on your screen that you are casting when you are "curating"
 
 
2 hours later…
1:51 PM
13 messages moved to Chimney
 
2:08 PM
eh, i think showing that you cast a downvote, would be an important feature of such a video/steam. Sure, it's deanonymizing your actions, but that's kinda the point.
presumably someone creating that kind of content would be willing to stand behind their actions and, when appropriate, admit when wrong and retract
 
"While"? To be fair, it looks like you perfectly acknowledge my point.
Let me try to be more clear: the network was born out of a site aimed at developers solving work problems mostly (or people who love coding). In that context the idea of "less talk, more content, cut the chatting, we don't care" works: people came to the site to find a solution to a problem and find it fast, not to make "friends" (wonder what, this is a recurring line you may notice "Not your friend/mind you business/I don't know you/etc" )
SO "average" user does not seek social interaction. So the "robot" approach work
As soon as you move to what in a way are - forgive the term here - fandoms... that model fails.
I don't expect someone going to Anime, Arqade, RPG, Blocks, Board&Card Games, Sci&Fi etc looking for that same "workplace" experience.
Especially when there is plenty of other platforms that offer a social one
Do you expect someone with a question on the last anime to go to Anime.SE to be told that their question is poorly written, does not need greetings and get no answer after a week or go to a Discord anime channel and talk with other fans there?
Short version: the company always tried to apply SO model to everything. It does not work.
And the chat, that "second level citizen" for some sites would actually be the main entry point, not the contrary.
 
2:26 PM
@SPArcheon that chat's underappreciated is... brought up a lot
 
2:41 PM
@SPArcheon I appreciate the clarification but I also would point out that the comment I replied to reads like ... well, to put it bluntly... it reads like how the company has been accused of blaming all of the problems of the platform on the "mean" users who rebuff newbies and gatekeep the site, pushing potential new community members away... rather than looking at how we've failed at alleviating that pressure on the community members to need to do the gatekeeping.
The main point in my response is to validate - yeah, I understand that's a problem but I think the core issue is not on the community to fix... if we don't educate/onboard so that people realize how SE is different, we're asking something nearly impossible from humans - to be perpetually cheery and kind when you've been worn down over the years by new people repeating the same mistakes time and time again.
 
eh, well, yeah i think that's absolutely correct, but, there's always going to be new people repeating the same mistakes time and time again. They're looking for personal help, a solution to their problem.
it'd be great if an on site search/summarization tool could provide for that problem, but also... there's a lot of people who seem to enjoy participating in that area
 
@SPArcheon I don't even agree with this. Or, at least, I'm going to question using "average" there - While the average user may use SO like wikipedia, which is fine... I think there's definitely a subset that would engage more if they could but we blockade everything behind privileges that they can't overcome unless they're willing to post on the site - you can't comment to get clarifications, use chat, vote... nothing.
 
answering those kinds of questions is also a fairly good learning opportunity for new users and new developers alike
if they aren't shut down by long term users coming in with their nteenth 8 paragraph expert answer that no new user/new dev feels they have the competency to compete with
 
I know it's not popular opinion but we need to move away from this whole posting == privileges thing. I've been really stressed for the last few weeks (months?) because I feel really strongly that this is an important change and I don't know how to help people understand that. There are risks but I think that the platform would be far better if we find a way to award privileges without tying them to reputation.
@KevinB I mean... not sure it's related but I definitely think there's this... unwillingness to accept "basic" questions... which I don't understand at all. If the question exists already and is answered in a digestible way, duplicate close it because the answer is there. If the question doesn't exist or the answer is buried in a 2-page treatise on a bigger subject, answer the f-ing question and move on with your day.
 
i'd like both of those things to be possible without it conflicting with curation
let people provide a personal specific answer to a given question, and let it be dupe closed, without interrupting the auto-redirecting for not-logged in users.
while somehow getting search to prioritize questions that have been used as dupe targets
but for some reason it's popular to throw downvotes at both
 
2:58 PM
I think maybe we have different things we're thinking about - since the stuff in my mind should be reasonable questions on their own. I think SO relies overly on canonicals in situations that don't make sense for the consumer of the content... at least not with the UI we have now.
If we'd built something like the answer labels that would allow people to differentiate answers by technology, that'd be different but scrolling through a two-page answer to find how to write part of a query in tsql because the answer covers all variations of SQL is... a lot.
 
i think that's a great example of what i mean
 
I have been fond of Tim Post's concept (somewhere) of "answering as a duplicate", where you can propose a duplicate and answer the question or cite an excerpt from one of the answers that answers the specific question the user has... it'd also be more helpful if we had page anchors within answers, even if only based on headings.
 
@Catija Correct. Minus the part of the blame. Seriously, let me tell you that any "bluntness" is not meant as "it is someone fault". What you see is resignation, not anger.
 
i don't want that tsql question with a canonical answer that covers it to not in some way reference that canonical, but also i don't see why we'd prevent someone from answering it too.
currently we want to prevent answers because valuable answers existing across a dozen questions creates a messy separation of useful information
 
That's fair, yeah. I don't think an actual dupe closure makes sense but I'd love to see something that connects individual questions to the canonical somehow so that people with the same question about a different variation of SQL can find the canonical more easily.
 
3:02 PM
@Catija that is the part I disagree with. I don't think that the model works for topics that are more shifted towards informal relationship instead of workplace formality.
 
@SPArcheon Oh, absolutely. Just because I work here doesn't mean I don't see (or even feel) the resignation. :)
 
i think there's a better solution to that separation/reptition than trying to prevent the flow of answers, rather, let the answers get posted and lower the thresholds on what gets deemed "not useful" enough to be deleted by roomba after a year or so
 
snort Y'all are reminding me about the question I've been procrasting about cause it feels too simple :D
 
@Catija Sorry, average was a poor choice. Make that "target"
Because that complies with the original motto too.
 
@KevinB I'd argue (not naming names) that tenure and subject matter competency are not related
 
3:04 PM
@SPArcheon Eh, I don't necessarily know that we disagree... I honestly didn't even get into the rest of the network in general in my discussion... and I'd say that if you know my background (and why I'm here), you'd have little doubt about how important building community bonds and having ways to get to know the people behind the posts is really important to me.
 
or somehow fix search so that the repition doesn't make it completely useless
 
there's a ton of stuff a 'new' user could school me on
 
(in all places we search!)
 
> Stack Overflow is a question and answer site for professional (and enthusiast programmers)
 
> and enthusiast
 
3:05 PM
SO is built for professional foremost, with an idea of a workplace-like format
Greetings? Not needed. Chat? Second citizen. Comments? You don't own my time, I have better to do. Friend? WHO KNOWS YOU CREEP.
not a social experience.
 
@KevinB I mean... the system isn't preventing or acting on the "usefulness" of a question... I'd argue that the current behavior of people voting and closing on SO actually indicates to me that we need a big overhaul of the definitions of what "useful" means. Which is why I really don't like "useful" as the word we rely upon in the voting hovers.
 
"Persons" are not the focus. Content is. You say it every day.
But that may keep professionals here and works for SO.
It doesn't for "social" passions.
 
@SPArcheon No. It's really not. SO was designed to be a library of answers to questions about programming. There's nothing in that definition that indicates those questions have to be "interesting" or "above some minimal level of difficulty". It's ALL - so this concept of closing questions that are "too basic" just pisses me off.
A 20 year python dev confronted with needing to use JavaScript is going to need to get answers to basic questions. Why should they not be able to get them on SO?
 
@Catija Keep reading, not what I meant. It is not about the complexity, it is how you ask. Alas, exactly what you said. It is a " library of answers to questions" that disregard everything else. It doesn't mater who asked the question, where they lived or what they liked. Only the question. Like you said before "It is not a forum".
What I am telling you is that a dev rarely seek out a documentation site because they want to chat.
But a gamer, or an anime lover, or a tabletop rpg player... they want to talk about their passion. Not post a question and be done with it.
That is the "engagement" the company miss.
So, if I want to ask something about the ending of my favorite anime...
Tell me: WHY I should come here **if not because I already have an account on SO (or another "tech" site)?
 
I don't think I'm understanding how the company is missing that engagement.
 
3:16 PM
poor wording again maybe, I mean that the company/network does not provide the type of experience / engagement I think fans of such topic want from a site.
 
I don't really see that as an issue? We have Chat (assuming you have the reputation) but the site is a Q&A platform. People want to be a part of narrow communities, generally and our sites are largely too broad to be a good place for people with very narrow interests. Why would you try to have a discussion about the series finale of your favorite anime on a site where there may not be anyone else who cares about it? Surely it's better off on the subreddit or discord where all of those fans are?
 
@Catija True, but then you can't expect some topics to really have that much success.
Obviously this depends on what you consider success.
Arqade is considered one of the historical sites. I see it as going quite poorly IMHO.
Anime is even worse.
 
We've been repeatedly reframing "success" on SE since before I was hired... in the time I've been here, we've moved from a rigid system where a site needed 10 QPD to "Graduate" to a recognition that sites can be small and still successful.
 
I don't have the data to really argue on this at all. I don't know what the "profit" target is, if any at all so I can't judge.
Yet this feels quite empty.
 
@SPArcheon Arqade is in a tough spot... I'd say Anime less so but still... Arqade is fighting with reddit, Discord, Discourse, and classic forums for attention and most people don't want a site that covers all games, they want to be where people who only care about their game are. - and to be clear, SO has the same issue, it's why we're poking at Collectives at all. People don't want the noise of stuff they don't care about. It pushes them away from the site.
Going to Arqade and seeing all Minecraft all the time was really annoying. I even play Minecraft now with my sons and I still don't like how the games I play don't even have one question, let alone the ability to answer all of them... and a lot of that depends on the games. We've moved into a space where games are changing weekly so a static platform like SE becomes less useful than it was when games were largely on consoles or computers and didn't change much after release.
By the same token, even as someone comfortable with SE, I don't want to start seeding content for a game no one has ever asked about on Arqade - why bother. No one's going to find it there anyway and I wouldn't want people judging my depth of knowledge of those merge games I play as a "guilty pleasure". :P I'd rather hide on their reddit or in their discord.
 
3:30 PM
@Catija You partially proved what I said a moment ago here, with the only difference that I would not really mind the "exclusivity" issue (But I recognize that as important too). Yet let me add that I thin Anime is in the same spot. Science&Fiction may too (especially against serie-specific channel like you suggested)
@Catija Let me try an example. I don't know if back in N64 times you heard about the whole L IS REAL thing - people trying to find a meaning to a texture in the game that looked like it had a blurred message on it.
 
All of that said, even then I see things that show me a site like Arqade is really needed within those systems - One of the games I play has a Questions channel in Discord and they realized that FAQs needed to exist and be easily accessible and regularly updated. When I first joined the Discord, people would just re-answer the question every time it was asked... and that gets old after a while, so they created a new channel with FAQs that link to answers to the most commonly-asked questions.
 
That thing became viral on sites like GAMEFAQS. Even after they realized there was nothing to find, people continued to fill page after page of forum posts discussing theories, documenting findings and so on. They kept visiting the site for the chatting.
Here that things would have been shoot down by you usual gold hammer guy with 1 vote close power.
 
I'd also say the complexity level of some of these games now is so high that single aspects of the game can have an entire site dedicated to just that aspect and not touch the rest of the game at all. There's really no value in doing anything other than pointing them to the existing site because it has all of the info and is updated regularly to keep up with the game's weekly release schedule.
 
another valid point.
but then, how can Arqade compete with GAMEFAQS or even game specific wiki sites?
How can Anime compete with dedicated communities?
Again, why should someone come here to ask their Dragonball question instead of reddit/discord?
 
I mean... I don't have much patience for fan theories, myself. Most people seem to dig way deeper into things than I can possibly care about... Like... why bother looking for hidden meaning in that sort of situation? I understand people do, it just really doesn't make sense to me. I just tend to take things more at face value.
 
3:39 PM
@Catija because people like to talk about what they like. Simply as that. They like to rant about what they would have done different. And they like to find others that agree.
 
@SPArcheon So, part of why I don't think it's as much of an issue with Anime is that... (excluding retconning) the medium is fixed. No one is coming through and changing Ichigo's hair color to anything other than orange, so the need for perpetual updating of content is ... not really needed. With games, that used to be the case... but it's not any more. Excluding speculation, I think SE is a totally useful space for questions about Anime.
 
@Catija Gomu Gomu no Mi wants some word with you...
 
@SPArcheon That example is more understandable - I can happily complain about sub-optimal decisions... but trying to find hidden meanings in a texture is just not ever going to make sense to me. If it's not as obvious as the Hidden Mickeys at Disney, I'm going to just assume it's unintentional if I even notice it.
 
@Catija Jokes aside, either the question can be found out easily on any site without need for Anime or it is complex enough to be either unanswerable yet or hard to a point no one wants to waste energy on it when there are questions about "Ichigo hair color"
 
Ugh, I have no clue why I wrote black... probably had Goku in mind ... I definitely thought about orange.
 
3:45 PM
i've always kinda wanted to participate on other stacks, but the ones i see as potentially interesting have little or no people discussing the topics I'd want to discuss, and then when they do... it's... quetsions that are just... iunno. dumb.
like i'm really interested in starfield atm, so i hopped over to arquade, and the interesting questions are self answered, while the rest are... things you could just open the game and there's the answer, why are you asking, kind of questions, so i'm not sure what i was even expecting
 
@SPArcheon So, yeah, the example is absurd, but if I want to know what the Bankai forms of all of the shinigami are in Bleach, that list is going to be pretty static. Will it also exist on the Bleach Wiki - sure... but most of the programming knowledge exists elsewhere, too... doesn't mean we stop trying to collect it. The issue is that many (most?) SE sites don't have the completeness of info necessary to fight for dominance against the subject-specific optoins.
@KevinB There's no U in Arqade :P Because it's arQAde :P
So, one of my favorite answers I've written - ever - is on Arqade but it's a weird one where someone was trying to solve a puzzle from a magazine. I'd never played the game but I suggested an answer based on what I could dig up around the internet and it went over really well... the thing is, I've never been able to confirm whether I was correct or not.
 
@Catija yep, and then "This question does not show any research effort"
 
Oh, good, someone brought up my other pet peeve.
It's an f-ing library. You don't refuse to have content in a library because it's in a library two states over.
 
Congratulations, @Catija, you have earned a new spell: Twilight Sparkle
^ don't worry if you don't get this one. @ShadowTheGPTWizard will
@Catija yep, shared sentiment here. Unless it is really "please read the manual"
 
I know people aren't close voting based on market share of views but I kinda wish they considered that. Yay, it's in a Wiki!... If it's in the wiki, it should probably also be on SE because otherwise, no one's going to come to SE instead of the wiki.
There are some exceptions, obviously. I'm a big proponent of the policy on M&TV to close questions asking who played a character in a known Movie/TV show. I will draw a line at replicating IMDb on SE.
 
4:01 PM
@Catija the irony is that most of the lost info on the internet comes not from wiki but from forums. Trust me, you can't even image the amount of lost info about pokemon when the Nintendo of America forums were shut down.
Because no one copied the info about that pokeball in emerald found in a room you could only access by cheating that would add a bad egg to your party when interacted with... and then save your game (something no other item in the game does)
I mean, most probably don't even remember what the "bad fish" was in Animal Crossing.
 
@DavidPostill I'm not a mod on any network site. not sure what PII I'd be able to leak at all except that which some user leaks themselves into the public space.
 
@starball well usernames count as PII, and I guess flag on posts ...
 
4:39 PM
@DavidPostill usernames are public info. I guess usernames on deleted stuff could be an issue with my 10k on SO. I can't see anyone's flags other than my own.
@Catija lmao. It's not super often, but I do somewhat regularly find myself linking people to meta.stackexchange.com/q/8724/997587 and meta.stackexchange.com/q/5280/997587 particularly when the definition of "research" comes up and whether it involves off-site research. it's why I asked meta.stackexchange.com/q/391467/997587.
 
@starball I mean... yes? But still, people have their questions downvoted for being "not useful" if they're too easy or they didn't "do research"... and it's just so confusing to me.
I love it when people throw around what SE should be and how votes should be used and also know that so many existing voters are voting based on what they define as useful rather than whether something is useful to someone generally.
 
@Catija the funny thing is that the easiest / simplest / most-basic questions are typically the most highest rated over time. and I've come to understand that that's just normal, and that it just demonstrates what "useful" actually means.
 
When you vote on a question, the core thing you should be asking is "Does the question contain enough information to be answerable?" If it's out of scope or whatever, that's what closing is for. Penalizing someone for asking a question that already exists (here or elsewhere) is... just not OK in my mind.
> HOW TF do I escape VIM????
 
@Catija in my understanding, sort of? like- containing enough information is part of the suggested upvote reason ("clear"). there are other conditions ("useful" and "well-researched" (on-site for the most-part))
 
@starball But new questions don't get that same treatment. I've seen new "easy" questions that don't seem to be duplicates get downvoted and deleted. Not that I can point to any specific examples at the moment.
 
4:54 PM
but the suggested upvote reason is a union of all three reasons (because of "and", you'd need to meet all three conditions if you want to follow the suggestion as some sort of law)
 
@starball But "useful" is too subjective and "well-researched" is just wrong. It's interpreted to mean "the user searched for an answer before asking the question" but what it's designed to mean is more like "includes all of the needed details"...
 
@Catija I've come to also not really get that. I can understand it because I used to think questions that were dead-simple to me were dumb. but I'm coming to the side of "what is obvious to me as an SME is not obvious to others", and also, answering easy questions is a great way to make rep and keep making it in the future :P
 
"Clear" doesn't mean that - clear is supposed to mean that you can understand what the question actually is.
 
@ElementsinSpace That's good news. I will start being more active on this site. like anyone cares lol
 
@Catija is that really what it was designed to mean? I thought that would be more covered by close-reasons like "needs details/clarity" or "needs focus".
modification to my above comment: answering "beginner" questions that should not be closed is a great way to make rep and keep making it in the future
 
5:00 PM
So... this is controversial but (while I won't advocate for it) I find it very difficult to find reasons to downvote a question on main sites if we consider that downvotes aren't super close votes. If a question contains enough detail to be answerable, is clearly asked so others can understand it, and is in scope for the site, it should be upvoted. Otherwise, it needs to be close voted.
 
@Catija ironically, it is not clear to me exactly what "understand what the question is" means
@Catija hey! who are you to tell me I should upvote something :P I'm a scrooge
@Catija but if you think in terms of site mechanics, downvotes are often more powerful even than a "super-close-vote": if a question gets a downvote and the score stays negative, and the conditions for roomba aren't broken, then the downvote is technically a "super-delete-vote".
 
@starball People have weird ways of explaining their problems and sometimes it's just not possible to understand what they actually want someone to solve. Hopefully that shouldn't be due to a language barrier but it can be. The end result is "do I actually know what question you're asking?"
@starball Absolutely. But that's solvable. We shouldn't be relying on downvotes to make deleting possible... it incentivises downvoting purely to get something deleted... which is why a lot of duplicates get downvoted.
 
@Catija still not crystal clear to me. often times I can understand a question to a large extent, but if I want to give a definitive answer instead of posting as an answer a big list of troubleshooting steps (as chatgpt so loves to give), I need more detail from the asker.
before I get that info to narrow down the problem and give a definitive, specific answer, I don't know exactly the cause of the problem, and in that sense, to me, I do not know "which question" (for which cause) I am being asked
 
Sure... but that's fine. Don't upvote. We have this dichotomy between up and down voting but it leads to people forgetting that not voting at all is also an option.
 
@Catija weird... I'm pretty sure the average user with an account forgets to vote instead of forgetting not to vote.
 
5:08 PM
If you need more detail ask for it - if they add it, awesome but downvoting while asking for more detail is disincentivizing them from improving the question. They see the negative number and they just decide it's not worth it.
@starball Oh, on average, sure. I was responding to your statement specifically, which seemed to be implying that you were trying to understand what to do if the question wasn't perfectly clear.
I think if I was going to rewrite the hover text for voting on main sites, I'd have upvoting be something like - "The question contains necessary detail, is in scope, and understandable." and downvoting would be very different - but not necessarily the inverse because that's what closing is for.
 
@Catija I think I get it (as in- I can sympathize). when I asked my first question and watched the downvotes come in, I can only describe the feeling as a mild form of despair.
 
Oh, and also, if you need more detail, you can feel free to close vote - I'm just not convinced that downvotes are helpful.
 
but it's also a hard reality for me to accept when my suspicion is that most askers have not read How to Ask
 
So? Are you voting on the post or the user?
 
@Catija the post.
 
5:16 PM
The post can't read how to ask.
 
I don't vote based on speculation of whether someone was read a Help Center page.
 
The entire discussion came from me opining about how terrible we are at onboarding users. They shouldn't be having to dig up a help center page before asking a question... we should be force feeding it to them and not letting them ask until they've shown they understand. We should be giving them templates to ask based on the type of question they have rather than just a free-form box and some generic text.
@starball Good - if so, why did you say the other thing? That's what led me to question it. And also, even if you don't, I'd bet others do. :(
 
@Catija I am both spectacularly astounded/shocked and incredibly pleased to hear a CM say that. I've been mulling over similar ideas.
 
If you go to skeptics you should find a quite long question about Ultima.
Went there, lost time writing it, got it closed, deleted it, deleted the account.
And.. I know how the site work. Imagine a new user experience...
 
@starball That's... surprising. I feel like I've been very transparent about my distaste for our onboarding and how many questions we let through that should have been prevented. But I suppose a lot of that has been internal. When we started the AI stuff internally, I pushed very hard to focus on search rather than an AI tool to draft questions and I explicitly stated that if we do it well, we should expect (and want) a drop in QPD.
 
5:24 PM
@Catija because I was thinking about systematic onboarding. You mentioned encouraging someone to provide more detail without discouraging them with downvotes. That led me to think about what the golden path for a question on SO is: ideally, if the asker follows How to Ask, that should rule out a very significant amount of cases of missing info.
It's hard for me to accept that we're so frustrated about discouraging askers when I feel like we aren't encouraging them enough and equipping them enough to take the happy path.
that's the meaning behind that comment of mine.
@Catija I need a cloning machine. make more catijas. unleash them on SO Inc.
 
... so there's two sides. We should be discouraging askers before they ask by shoving existing questions in their faces to reduce the need for new questions. If that doesn't get them a solution, we should have a full question-asking workflow that supports them to create complete questions. The Question Wizard is the closest we've gotten to that but I think V1 was better in some important ways than V2... so I think there's a lot of room for improvement.
Once the question is asked, it should be supported by the curator community to make any outstanding improvements that the UI couldn't predict or whatever but the reduction in questions overall should alleviate a lot of the pressure on curators so that they have more time with each question.
 
@Catija this is where I insert my gripe that the possible duplicates section of the Ask Question interface is (IMO) ridiculously small (in terms of height) and that its shortness actually seriously discourages me from using it. mouse scroll sensitivity and granularity is a thing. When I scroll, I often lose where I am in there (one scroll increment of my wheel and I've skipped over entries).
literally I don't want to use it because it's not a good user experience. to me, that's a and frustrating and stupid problem to look at because it's so fixable.
 
Yeah... but you shouldn't be shown duplicate candidates in the asking flow, you should have already gone through a pre-drafting step that finds the questions for you.
 
@Catija I floated a related idea in here recently: chat.meta.stackexchange.com/transcript/message/9642091#9642091
 
My favorite part of V1 of the question wizard was that it had honey pots that'd send you to other SE sites if you said you were having a hardware problem or were looking for a software recommendation :)
 
5:36 PM
@Catija Was V2 the one somepony had repurposed to write MLP short stories?
 
and then revoke the badge and let them take the quiz again if it seems like they could use a refresher (Ex. if a question of theirs gets closed)
@Catija going back to this idea, the similar idea I was mulling over was less dramatic (not-force-feeding). basically, giving each Help Center page about asking questions a very basic comprehension quiz, and associating a bronze badge with the quiz. then letting answerers filter for questions asked by people who have those badges (basically a time-saving mechanism for people who don't have a lot of time and want a higher chance of questions that are answerable).
 
5:52 PM
@starball Sure, even something like that would be... something... but you have to funnel them into the article somehow or they won't even go looking for it or know they need to read it. May as well just build it into the asking workflow.
@starball did these get flipped? Regardless, yes. As I work to separate rep from privilege, I want to ensure that we're considering how to revoke privileges. For example, a user who is robo-reviewing can be review banned but they retain the close voting privileges perpetually, even then. Why???
 
6:43 PM
@Catija lol. well, there's a badge for review, but not for close-voting in general (and intentionally like that for pretty much that reason from what I recall reading.) assuming the badge is what motivates robo-reviewing.
@Catija I was trying to think of something not too too groundbreaking to increase the likelihood of it being something the company would consider implementing. and even then I was afraid of it being too much dev work to be considered by the company. so that's why I was shocked to hear it coming from someone in the company (granted, you are a CM).
I've also heard from Yaakov in the Staging Ground chatroom that they have similar ideas about systematic onboarding for being a Staging Ground reviewer, and hopefully extending it (system for systematic onboarding / training to use privileges) to outside Staging Ground too.
really great to hear that people in the company are pushing for this kind of thing.
 
7:09 PM
@starball I think my concern is that I've seen people who intentionally close vote/flag/review based on their personal opinions of the site scope and the only way to address that now is a total suspension. That seems... ver heavy-handed. I don't like preventing access generally when only one behavior is problematic. I would rather make it possible to revoke some privileges if it seems necessary to address behavior.
 
@Catija yeah. the hard part is that a lot of info or finer detail about when and when not to close-vote and flag is not very discoverable. either in the Help Center (which itself is not as discoverable as I think it could be), or in a meta FAQ, which is also not very discoverable, or in meta discussion, and sometimes scattered across multiple meta discussions and sometimes even worse, multiple meta discussions with different consensus, or changing consensus over time.
 
Even when it's not intentional, we just give anyone with 3k rep equal access to closing with no indication that their knowledge of site scope is correct or current. Someone coming back after several years away could unintentionally be unaware of policy changes. By requiring some number of close votes/reviews per year to retain the privilege without a refresher or something, we ensure the curation is currently and more consistent.
 
(continuing my previous comment) which also highlights that to an extent, when to close and when not to close is a matter of "opinion", but group "opinion" (consensus) about topicality of a site that comes from meta discussion.
 
@starball my big idea is to have a default set of information about the network wide close reasons with site-specific information created by mods (or at least input by them) so that it's customized per site and kept up-to-date. We have pretty good explanations for the close reasons, they just need to be more discoverable.
 
@Catija SO's FAQ is here. Also not very discoverable. I didn't even know that MSO had a FAQ until years after I joined the network, and I'm pretty sure I just stumbled upon it while not actually looking for it.
has a section on not voting to close for lack of effort
 
7:18 PM
@starball This is why I love the new Close UI and how it customizes the guidance based on where it's shown and who's seeing it. I'm really frustrated that I haven't followed through on updating the network wide reasons but I was struggling to finalize the text based on feedback to the meta post and I still need to get dev time to actually make the copy changes.
 
though it will probably bother you greatly that it says this: "If the question is lacking in effort or research but is understandable and answerable, consider downvoting and / or leaving a comment instead of voting or flagging for closure."
I do think context matters though. I don't really think about effort when voting, but I do think about whether I think something is likely to have long-term usefulness by how much I think someone else will have the same question in the future.
 
Ugh. Yeah. I have some thoughts I'd like to share on MSO at some point about how they're handling stuff but I'm also trying to avoid looking like a Company voice coming in and telling them that they're off the rails in how they're voting and closing questions. 🀣
 
actually speaking of effort, when I see a "how to" question, I usually don't like it when I see a failed attempt. for one thing, if someone just wants to know how to do something, and the question isn't too broad, nobody in the future will care about that wrong attempt when reading that Q&A. All they will care about is how to do it right.
And when I see a wrong attempt in a how question, I feel compelled to explain why the failed attempt doesn't work. The wrong attempt and explanation of the wrong attempt I think are likely just noise / a distraction to future readers.
 
@starball long term usefulness is... a consideration but I don't know... there's a reason we got rid of the "too localized" close reason. Code review actually asked whether we can remove duplicate closures entirely since they have such a different format than any other site. Everything for them would be unlikely to have lasting value... but sometimes I think we're a bit too concerned with that, too. Not to the same degree as other things, but somewhat.
@starball interesting. I wonder how you feel about my SU question, then. I intentionally included as much detail about what I've already tried for two reasons - I didn't want the question closed as lacking detail or effort and because I didn't want to get answers that told me to do things I already did. That said, I linked to the solutions I'd tried so that the question isn't totally useless for people who might need those solutions.
 
@Catija lol. the CM position sounds like a difficult one to carry / "wield"
 
7:31 PM
It's harder for some. Shog never seemed bothered by it but people trusted him implicitly and had no reason to believe he was doing anything other than trying to get the ship on course or show how it's diverged. He also has tons of experience on SO and as a programmer, which gives more weight to what he says. I rely more on my brain and its obsession with making systems functional and frustrations when they do not.
 
@Catija I was talking specifically about "how-to" questions and not about questions about diagnosing/identifying problems. In questions about undesirable behaviour that one is not exactly sure the cause of, giving as much detail as you can about diagnostics is good, and I wouldn't classify that info as "failed attempts". that info is valuable because it saves people suggesting what you already know does not help solve the problem.
 
Because of that, I have to be good at accepting where my proposal won't fit into the reality of the system and adjust accordingly. A good system should support everyone who wants to be in the system to do what they want to do while being cognizant of people who want to misuse the system. It's necessary to find ways to prevent the bad without impeding the good.
 
I'm no stranger to questions about undesirable behaviours where the asker is not sure of the cause. I get many questions like that in the VS Code tag. Very many of them are resolved by going through some pretty simple troubleshooting steps, to the point that I built a browser extension just to help me post comments asking for info from following those troubleshooting steps
 
@starball ah. I see. I'm not a programmer but if I were asking how to get my frosting color more saturated, you'd prefer I didn't explain that I'd already tried gel-based colors instead of water-based ones? 😁
 
and in many cases, for questions about undesirable behaviours that the asker doesn't know the cause of, that kind of diagnostic / troubleshooting info is essential to giving a answer specific to that problem instead of answering with a general list of troubleshooting steps (which I already ranted a bit about above)
 
7:35 PM
Or is that still more on the troubleshooting side of things?
Maybe an example how to question would help. Would you classify them as maybe a "best practice" sort of thing?
 
@Catija I'm not much of an arts-and-crafts-er :P maybe (probably) I'm overgeneralizing / speaking for a specific context. for example, I spend a great deal of time explaining why an asker's failed attempts don't work in a question that is really just a how-to question here: stackoverflow.com/a/77205819/11107541
I didn't need to explain the failed attempts, but as I said before, I often feel compelled to.
 
@starball I totally agree but it can end up seeming "too localized"... in my case, if someone doesn't have third party cookies blocked, my answer won't help them.
 
The other option is to remove the failed attempt from the question, but that just doesn't feel right (violating the author intent guildeline/rule)
 
My brain keeps reading CMake as cake, so it's not making a lot of sense but I see how you were explaining why their attempts to solve can muddy the actual answer.
 
@Catija ideally, the Q&A would just contain exactly the information which would be needed to scope answers to the specific problem. one of the principles of self-answering is that you don't get a free pass on detail and clarity in the question when you self-answer. the question should ideally be able to rule out answers that are not about the problem that motivated the question that was written. of course- the asker often doesn't know which info is relevant and which is not (though an SME can).
so another option for me when I see a how-to-question with a failed attempt is to write up the question and answer it without the failed attempt. I haven't done that much. can't recall right now if I've ever done that actually.
 
7:42 PM
yeah.... iunno. i feel like upvotes should be for cases where you literally needed what the post is providing. Not just you read the post and it has a valid question and isn't closable, etc,
put another way i don't think it should be a participation trophy, it should be a measure it being proven valuable
 
@starball Eh, I think removing extraneous information, even if it's not random "thanks" or whatever, still fits well into the concept of keeping questions concise and reusable. I think as someone answering, you have a few options... personally when I'm in that situation, I answer the core question first so that it's easy to find the answer and then I will put a horizontal rule and add some explanations about why their solution was not working.
 
unfortunately questions rarely carry value, the value they bring is in attracting answers.
 
I think it's fine to remove it in some cases, though, simply because it overcomplicates the question and can confuse future readers if they think it might be a viable solution. Authorial intent must be balanced with the need of future readers. I'm a big fan of "heroic edits" on small sites where you just rewrite the question into something useful.
@KevinB I feel that way about answers... I vote rarely because I don't often know for certain that the answer solves the problem... I know sometimes I just know the answer is correct because I have knowledge of the subject and have used the solution before but I think a lot of people vote when an answer is plausible rather than true because they don't have the time/interest to confirm it works.
 
okay I'm digging up instances to self-expose myself when I write up a question in response to a question that I don't feel comfortable changing into what I think is its ideal form:
 
I think with questions, upvotes can /should be used to indicate cases where the question is something I might have wondered in the past, even if I'm not currently struggling with the issue. If we limit it to the degree you indicate, someone answering would be unlikely to feel they can upvote the question unless the reason they're answering is because they self-solved. I don't think it's unreasonable to discourage voting on questions simply because you don't have the problem.
 
7:52 PM
@Catija let me give you an extreme example- a question which I initially wanted to close as too broad. take a look at the answer I ended up giving and you might see why: stackoverflow.com/q/21163188/11107541 there isn't really a failed attempt, but there's a lot of stuff in there that is a lot to unpack, and the question is deceptively short, whereas to unpack everytihng that it's about, it takes a lot
@Catija personally, I generally only upvote things that help me. If it doesn't help me, but I could see it helping someone else and it shouldn't be closed, I just don't vote, and answer if I can.
 
@starball what's extraneous in the first example? I'm not certain about the technology at all so I'm totally unable to even recognize whether the questions are equivalent.
@starball interesting. So even if you would say, "wow, I had this exact problem when I was first starting", you wouldn't upvote because it's not a problem now? Ugh. If that's the case for many people, I feel even more convinced that we need to remove the rep to vote. If no current privileges users have the question right now, the question will just never earn votes even if it's getting tons of views because it's a common issue.
 
@Catija sorry I didn't really explain. those aren't really about extraneous info. there are other reasons why I might not feel editing something into what I think is the ideal form.
 
i do think that's fair, for questions
 
I'm sitting at a gas station right now and I feel weird not moving for 20 minutes. I'm going to get gas and go home so I'm not pecking away on my phone at a gas pump.
 
@Catija yes, that is how I do :P
@Catija well, that's just me and how I vote. in my experience, the votes eventually come. especially for high-visibility issues. take a look at the questions for my top-scoring answers. many of them are for high-visibility bugs.
another example of a high visibility bug: stackoverflow.com/q/76247741/11107541
@Catija interesting... I haven't run into meta discussion on this (probably because I haven't really gone looking for it). I've just been scared to do these most of the time (scared that someone will someday come scold me)
 
8:04 PM
I think any discussion of voting has to be thoughtful about the realities of people. We can recommend how to vote and explain voting best practices but we can't read the minds of anyone voting or assume any specific reasons for the votes. I think it's fair to consider finding ways to normalize post scores without relying directly on votes, I just have no clue how to do that.
 
and also, to be honest, there's an aspect of "if I know how to make a much better version of this question (generalizing the problem as much as it can be generalized, and omitting any information that isn't necessary to properly scope the Q&A), and it would be significantly different than this person's question, why should this person get the credit for writing that better version of the question instead of me?"
for old questions with lots and lots of votes, then I don't bother (again, scared that if I try something, someone will come scolding me), but for newer questions I do think about that.
@Catija normalizing for what?
 
@starball search for heroic edits.
13
Q: Rep for heroic edits

RosinanteI'd like to present a case related to Let me reward a good edit on my question/answer. If you close this as a dup to that, ChΓ© Sera, Sirrah! I submit that the three most valuable contributions to the sites, in descending order, are: Good Answers 'Rescue' edits Good questions What do I mean b...

 
@Catija interesting
 
@starball any number of things... rep/expertise of the voters, when the votes were cast, how many votes the voter has cast, anonymous votes, views.. any number of things.
 
@Catija also interesting.
@Catija actually, I've been mulling over an answer post to the 1-rep voting post you made recently about interest in experimenting with visibility of anonymous votes.
 
8:12 PM
@starball it's a fair concern and I think it depends on the person. While getting credit is awesome, editing can potentially help a new asker see what a good question looks like without you seemingly "stealing" the question from them. We see very similar concerns with people hand wringing about converting a comment into an answer and not wanting to "steal" the credit.
@starball I clearly need to understand them better myself. Where I've looked at them, what Jon wrote in his answer seems to be the case... most of the votes are from logged-out users, particularly on the posts that get the most votes per month.
But it seems we only retain one vote attempt from logged out users per IP, so if multiple people on shared IPs attempt to vote, we don't know that, from what I can see.
 
@SPArcheon :-)
 
@Catija how does that work with upvoting and downvoting? say one person on the IP upvotes, and then another downvotes the same thing. is the last vote taken? the first one? the average value?
 
@starball no clue. It's something I was thinking about checking. Unfortunately we cache the votes so I'd probably have to test on separate days since I don't know how frequently we save to SEDE.
 
8:34 PM
@Catija you can't poke a dev to investigate? I guess they're too busy?
 
@starball nah. I can. I have access to GH, so I can technically look for it myself, I just don't know what to look for or how to read the code to understand it.
 
@Catija just add me and I can do it (maybe) :DDDDDDDDDDDDDDD
@cocomac hehe look at meta.stackexchange.com/a/393276/997587 "cocomat" I'm not fixing it xD
I now dub thee "coconut mattress"
 
I wish more of the code was Open Source. I feel like it would really make it easier for people to understand the mess but also how things are built. We'd have to be really careful about stuff that can't be public, though, like vote tooling and stuff.
2
 
@Catija not sure how feasible this is, but one idea would be to put functions and constants for anything you want to be hidden in a separate, closed-source library(s) / packages / whatever you call them in C# land
 
I think the opposite is more likely... and less risky? Sort of how Stacks and the new editor are open source... so we'd find high-value stuff to explicitly make public rather than trying to hide everything that can't be. But I'm not a dev or an architect ... I can barely use GH.
 
8:50 PM
@starball ;)
 
@starball Adam says we only store the most recent vote in SEDE but we have another tool that logs all vote attempts, though it's unclear whether that data is tied to specific posts or just a count of all "upmod" and "downmod" events.
That said, there's a bug in some views that doesn't make the vote indicator disappear after a vote fails for anon/unprivileged users, so we've seen a spike in the vote attempts per day that I'm guessing is due to the bug rather than to the buttons actually causing more people to vote. We're trying to see about fixing the bug and seeing what happens in the data.
But it's the weekend now and we all have Monday off.
🌴🌴🌴
(I've been off today, if anyone wonders why I've had so much time to chatter.)
 
9:09 PM
@Catija say thanks to adam for me, and thank you for poking a dev :)
@Catija you're a gem for spending your day off in SE chat
see- example of something I think has long-term value and wish people wouldn't downvote: stackoverflow.com/q/77242748/11107541 (pls no meta effect)
 
9:26 PM
 
@starball 😬 I appreciate that... though it's less me doing work when I'm off and more catching up with my hobby. I don't know that I have any interests deeper than how I feel about SE.
 
 
2 hours later…
11:23 PM
@Catija Sometimes... the CMake is a lie
 

« first day (4811 days earlier)      last day (151 days later) »