75 comments

[ 0.24 ms ] story [ 145 ms ] thread
Is this the consequence of having entitled moderators for years, only now instead of powering over random users they finally turned against their creators?
I don't know why you say "entitled moderators". If you send any time with them, you'll quickly see the vast majority of the job is drudgery. You wouldn't say "entitled janitors" would you?
(comment deleted)
I might say "entitled HOA board members". Their job should be (in theory) pure drudgery.

And yet they find it so thrilling when they're measuring the grass on your front lawn with a ruler even though they live in the same neighborhood as you and know that it's rained every day for the past two fucking weeks without even 24 hours in between for it to dry up enough to push the damned mower through it.

That's . . . not a bad analogy. Only instead of 100 homes (like the condo I lived in with an HOA) imagine thousands of people complaining about their neighbors each day. Sometimes it's the ticky tack lawn nonsense and sometimes it's a leaky water heater flooding the neighbor. (The second happened to my neighbor. Unfortunately, it was my water heater.) Until you understand the scale of what the SO moderators were doing, it's easy to call them "entitled".

But I know it's hard to drum up sympathy for mods (or HOA boards).

I feel everyone complaining about Stack Overflow should become a Stack Overflow moderator for a month. I am 100% convinced almost all of them will severely nuance their opinions after that, if not radically change them.
Admins routinely interfere in elections to make sure they get the "right kind of moderator". This has happened personally do me, on one of the other SEs.
Every single time I've seen this happen it's been for a good reason, but of course I can't comment on your specific case.
When Stack Overflow refers to "moderators" it is referring to the people who have a diamond after their name who were elected as part of a moderator election.

These individuals are mostly doing things like stoping fights in comments, locking (not closing) posts that have debates on them, and investigating vote collusion so that people can't boost their rep from an alt (or coworker or spouse). On smaller sites (not everything is Stack Overflow) moderators are often involved in guiding new users as you can read all the posts from a day in a few minutes - for example Academia gets about 7 questions per day (Stack Overflow gets that many each minute). They also often run anti-spam tools.

Most actions that close questions (especially as duplicates), edit posts, ask questions in comments are from other community users (not moderators) who have rep from contributing to the site. Again, these are not moderators. Some members of the community have more of an interest to curate it to fit their ideal of what the site should look like (and that's a debate that goes back a long way).

It is the diamond moderatos who are going on strike. Some of the community are also further abstaining from doing things (answering questions, closing questions, doing reviews).

So the claim that it is the moderators that are going through and checking things and insisting that they are of the right form is misplaced. Moderators are rarely involved in dealing with the quality or content of a post unless that is brought to their attention by other users and there is a substantial problem with it (e.g. the person is posting with every variable name as a vulgarity and rolling back any attempts to edit it).

SO mods can sometimes be just as petty and power-abusive as reddit mods, which is annoying to see, and of course there is little recourse.
The recourse for a power-abusive SO diamond mod is to contact Stack Overflow the company through the contact us form and explain your grievance.

The number of moderators on Stack Overflow (and the rest of Stack Exchange) is significantly less than those on Reddit and the standards that they are held to are much higher, the tools that they are given are much more powerful, and the expectations of their conduct is also much higher than Reddit.

Their interest in correcting your grievance is directly proportional to your ability to cause them trouble if it remains uncorrected.

I have zero capability of causing them real trouble, and they therefor have zero interest in correcting anything.

Likely, few people have non-zero capability when it comes to causing them trouble.

> and the standards that they are held to are much higher,

I've seen no evidence of that.

The moderator agreement is viewable at https://stackoverflow.com/legal/moderator-agreement . There is nothing equivalent on Reddit.

If you see a diamond moderator violating anything described in that contact Stack Overflow (the company). There are serious repercussions for a moderator violating that.

Note that this only applies to moderators - those users who have a diamond following their name. Other users are not moderators (and this is one of the sources of frequent confusion - terming anyone who can vote to close a question as a moderator).

Well..

If a janitor were to come by my desk mid day and say. "That's not how you use a trash can. You must follow a prescribed format for putting things in the trash. I have decided that you are not allowed to use the trash can for 10 days. If I see that you have a co-worker, and they attempt to put trash in like you, I will also restrict their trash privileges for 10 days."

Then yes I would call the entitled.

No matter how drudgy a job is. If you squash people learning, because they can't live up to your standards, you're entitled.

Well your question was apparently so uniquely bad that they recognized it and correctly identified the duplicate by your coworker. So maybe you're just salty because they were right.
Stack Overflow was never intended to solve every single programming question out there, or to help every single person trying to learn.

You also can't show up to ballet lessons and start swing dancing.

I used to use SO all the time while coding. Now I use GPT-4 instead.

Those of you who anticipate continuing to use SO: why?

Because I want to learn to fish, not have fish delivered to me.

My last hit on SO was a search for “.gitignore wildcard”. What’s GTP-4 going to do with a two word prompt?

I agree with the first sentence, but unfortunately ChatGPT does a pretty good job with that particular two word prompt.
Made me curious, so I tried that exact prompt in the GPT-4 playground. Response:

  A wildcard in .gitignore file is a special character used to match multiple file names or patterns. The most commonly used wildcards are:

  1. `*`: This wildcard matches any number of characters in a single path component. For example, `*.txt` would ignore all files with a .txt extension.

  2. `?`: This wildcard matches any one character in a single path component. For example, `file?.txt` would ignore files like "file1.txt" and "file2.txt" but not "file12.txt".

  3. `**`: This wildcard is used to match any number of characters across multiple path components. For example, `**/build/` would ignore all files and folders named "build" at any level in the project structure.

  Examples:
  - `*.log`: Will ignore all files with a .log extension
  - `**/.DS_Store`: Will ignore all .DS_Store files in every folder
  - `build/`: Will ignore the "build" folder on top level directory
  - `/*/temp/`: Will ignore any "temp" folder at the second level of the directory structure (ignores sibling/temp, parent/child/temp but not sibling/child/temp)

  For more information, you can refer to the official git documentation on gitignore: https://git-scm.com/docs/gitignore
This is slightly wrong. build/ will ignore the build directory at every level. /build will ignore the build directory at the top level.

I also have no idea what it’s trying to communicate with its /*/temp/ examples

> This is slightly wrong.

I wonder how it does if you tell it it’s wrong. I’ve tried a few times and it seems to get confused pretty easily.

"I apologize for the confusion. [Repeats the same wrong thing]"
> This is slightly wrong.

A perfect representation of my git usage then!

I once had a project with a "bin-name" in the gitignore, to exclude the "/bin-name" that "go build" will produce. Then I added a "dir/bin-name/file.go" which of course got excluded. The CI/CD builds failed, but it worked locally, and great confusion ensued. Since then, I've always added that leading "/" and generally try to be as specific as possible in gitignore files.

This is a great example where many people providing different answers and comments based on their different knowledge and experience provide value, opposed to just ChatGPT telling you stuff. Simply "correct" is nice, but "correct, complete, and with experience and context" is better.

A priori, you can’t know whether this is accurate or a hallucination.
I really think the whole "hallucination" thing is a terrible framing. All that gpt-4 does is spew out bullshit. Now, it applies a bunch of statistics to attempt to spew out believable bullshit and those statistics cause it to sometimes accidentally be correct. The base assumption with this should be that it's wrong, and so we should instead be framing it as like sometimes accidentally producing "actualities" or something.
Ironic, since even if you're using SO, you're simply having the fish delivered to you in the form of answers to your questions there. If you really want to learn to fish, read the documentation and code more to build up one's skill.
I read both, the offical docs were the second hit. People who want one answer to everything scare me far more than AI.

I don’t really accept that SO is just answering questions for you. There are usually multiple answers. There were several comments on SO that went into more depth than the official docs. There are second opinions, and answers pointing out what the wrong answer is. There are links to related questions.

It wasn’t what I was looking looking for initially, but sometimes answers to questions you didn’t ask are better than ones you did.

I just had this experience today, trying to decipher the output of an obscure linux command argument, failing to dig it out of the man page; Google sent me through almost related stack overflow pages. Then I put a question in Phind and it gives me all the details, beautifully formatted, just a little too verbosely but with all the main points covered, with references.

It's time to update my search reflexes.

It's the opposite for me, I tried Phind a few times and couldn't get a valid answer to questions I could answer myself... Perplexing.
Isn't that like, exactly the same thing? - because GPT-4 was likely trained on SO (and reddit) data.
(comment deleted)
SO gives me better insight into who provided the information. When an answer is provided by a project’s main developer or a contributor I know I can rely on the info. From what I’ve seen with GPT it’s a popularity contest. Common mistakes and misinformation are regurgitated like facts and the high quality grammar + spelling makes things sound plausible.

Ask it about something you know is a frequently repeated mistake.

Those of you who anticipate continuing to use ChatGPT to answer programming questions: where do you think it'll get answers from if people who actually understand stop using q/a sites?

Where do you think ChatGPT finds these answers?

Same place copilot does, github.
Idk how often I've deleted my account there. Super toxic place. Some guy has a bad day and can act like an asshole because he has more points and knows people there.

SO gives people with points the power to bully others. When you protest you get collectively downvoted to the point that you a temporary banned, for asking a valid question.

For instance the last interaction there, someone closed my question as duplicate, but it wasn't a duplicate, so I asked in Meta how I can find out who closed my question, because I wanted to explain to them why it's not a duplicate. That question on meta got 30+ downvotes and the question on SO got them as well including passive aggressive comments like "do you know the answer now".

What is this 3rd grade mentality?

And because of that people use alt accounts to farm points. They use other accounts to ask questions, and answer them with their main account. Or let their buddies ask and they answer and the other way around.

SO is a terrible place nowadays. It used to be a very good place a very long time ago, had its ups and downs, but nowadays it's just a toxic shitfest

That's my impression exactly. I was thinking "junior high" rather than "3rd grade" but yeah.

Hollywood could get a good script for a sequel to Mean Girls just by watching those people.

I’ve never had much of a problem. However, I occasionally find a question that matches my problem exactly, but closed as a duplicate with a link to a question that’s similar, but not the same :-(
This is why a major value prop for ChatGPT is it's like Stack Overflow but it will always be nice to you :)
ChatGPT is mostly wrong though, so not helpful if you actually want to solve a problem. It's entertaining though
This is not true. I've been building a site with Sveltekit and Appwrite for fun. I know nothing about these technologies. ChatGPT has saved me hours of messing around and I haven't needed SO once. It doesn't always give me the best answer first, sure. It's an iterative process, which all software development is anyway.
> ChatGPT is mostly wrong though

Just like SO then

That is mostly correct but thing is, if you go on stackoverflow or chatGPT to have someone else write the solution for you, that's on you to use the wrong solution.

I've used chatGPT (as well as SO as many others) but always in the context of having enough knowledge (either from experience or experimenting before opening a question) to understand if the answers were just wrong or not.

To each their own but I treat ChatGPT as a rubber duck 2.0 and I think that's what it excel at.

This is why AI is going to take over the world in the future. Why bother hanging around real humans, who are frequently assholes, when you can hang around AIs that are always courteous and helpful? And once they develop AI-driven companion robots that look and feel mostly human, romantic relationships between humans will be gone: why bother risking a toxic relationship when you can spend your time with an AI that's always nice to you?
> SO gives people with points the power to bully others. When you protest you get collectively downvoted to the point that you a temporary banned, for asking a valid question.

If this happens, there is no way to recover. That question is forever blacklisted, not just closed. No one could ever ask it again. For a particularly long term problem I was having, I asked for a friend to repost it with his account. He wrote up the question, the verbiage wasn't mine (let along copypasta). I'm not sure what gave it away, but his version got -4 within an hour.

There are no effective appeals, no oversight, no reform possible.

And, if we evaluate it honestly what they've strived for the last couple of years, they don't even want the site to be a place for people like us to ask questions or answer them. They see it as some Wikipedia-like encyclopedia of "good questions". If your question isn't a good question, it does not matter that you need answers. It doesn't matter that other people out there have the same sort of question.

It's just not something they want.

Does anyone remember back in 2007 or so when we were all excited that now every link wasn't Expert Sexchange, and we thought the internet and the world were slowly being improved?

> They see it as some Wikipedia-like encyclopedia of "good questions". If your question isn't a good question, it does not matter that you need answers.

An encyclopedia of good questions is a very valuable thing. SO was created with the goal of it scaling, because once you answer the good questions, nobody needs to ask anything anymore.

Then they decided to write a formal specification for "good question". This is when things started to break.

Yep. 100%. Before good questions would be upvoted. Bad questions would be down voted. Get to zero and the question remained. The dev could revive it with a few up votes.

The problem is the gate keepers, seem to hover over the questions queue. I shudder to think how someone who has english as a second language would fair.

Noobs don't know how to form questions. That's what we're supposed to be here for. To help them learn answers, and how to ask better questions.

> The problem is the gate keepers, seem to hover over the questions queue.

I would encourage you to look at https://stackoverflow.com/review and the participation in those queues.

If you believe that people are gatekeeping those queues and preventing good questions from getting out, help out in the queues and work to fix up the questions too.

As moderators are on strike and things are not going through the queues, this would be an opportunity to help out the rest of the community.

I will point out that typically the queues have orders of magnitude more things needing than people can do in a day - individuals are limited to how many reviews they can do in a day so that they don't burn out or get in the habit of hitting the same button again and again.

> because once you answer the good questions, nobody needs to ask anything anymore.

If that's your belief, then nothing I could say will ever convert you away from such a peculiar religion.

(For everyone else) Not only will many need to ask questions still, but the "encyclopedia of good questions" is rather useless to many of us. Some small fraction of our needs is about a classic textbook example like out of some intro class. We live in the real world.

The "encyclopedia of good questions" has more to do about soothing the egos of jackasses who are (somehow) ashamed of the great thing they accidentally created.

> so I asked in Meta how I can find out who closed my question, because I wanted to explain to them why it's not a duplicate.

you're supposed to edit your post to explain why it's not a duplicate (and actually, before you post, you're supposed to search for existing Q&A, and when you post, to show what you have researched). In the duplicate closure notice, there's a big button suggesting you to edit your post if it's not a duplicate.

> And because of that people use alt accounts to farm points. They use other accounts to ask questions, and answer them with their main account.

This is called fraud and it's not allowed and seriously looked down upon by the community. The system has mechanisms in place to auto-detect fraud and ding people for it.

Nah doesn't work. I've had the same thing. I've even linked the similar question in the original, saying how I needed more. Editing the post doesn't bring it back to life. It stays dead. Like pulling the knife out of a stabbed corpse. They are still dead.

I've seen it a pile of times.

Back in the day, questions were all valid. It was the way devs should work. If you posted an identical question, so what? Who did it hurt?

This "fraud" is totally allowed. Who looks down on it? I could give a snot about my points. If someone wants to have 10 million "magic points" I don't care. Every dev I know, with 20+ years in has about the same < 5k points. We don't care.

I gave one dev 2k points for answering a single question.

Posting duplicate questions hurts the community. The whole idea is to have one answer for one question. If the answer changes over time it is updated. This does not work if there are 500 identical questions.

SO is not your personal helpdesk. It is a Q&A site. If a question has been asked and answered it is up to you to find it and use it, you don't just ask the same question again when people have already taken the time to write a good answer. That would defy the entire point. Maybe you should try asking on Reddit or a Discord server or some other community more appropriate for the kind of interaction you seem to want.

There are so many nuance to a problem that the approach of "one question - one answer" is simply naïve.

Even if we wanted to go with this approach, questions that are potential duplicate should just have an answer pointing to the other question and let the OP decide if it's right or not.

Having someone else that doesn't have the full vision on the question and context is not the right approach.

You can't let the OP decide if their question is good. Every OP thinks their question is good and most of them are idiots. I guarantee you most of the people complaining about SO do so because they suck at asking questions.
As someone with a 9-year-old account on SO, I have never asked a question there because I didn't fancy the chance of it being immediately shut down.

From personal experience, I have come across numerous questions that appeared to address the exact issue I was struggling with, only to find them closed as duplicates, off-topic, or something similar. Often, the suggested alternative Q&As are unrelated except for the surface-level similarity, which is especially frustrating.

Sometimes, when I find one of these closed questions, I feel a strong sense of empathy towards the asker, as if I'm the only person out there who understands them.

Unfortunately, I have always been primarily occupied with whatever problems I was dealing with at the time, so I don't have a list of examples I can share with you.

SO is undoubtedly a valuable resource, and I don't completely disagree with the moderation style. However, it has been a slow burn for me.

I don't doubt that this happens. Mods are people and people make mistakes.

Though i also think it is in part that the Asker doesn't understand how to apply the answer to their issue rather than the answer not being relevant. In my experience a lot of people are unwilling or incapable of adapting an answer to their situation, which can be problematic in a place like SO that doesn't attempt to cater directly to individuals like that but rather to answer more generally.

I think issues like this contribute a lot to the dissatisfaction we see surrounding SO.

> and most of them are idiots

Change the word most for some and you get a statement that also applies to mods.

I like to check SO for problems I encounter but no way I would ask a question in there. Really don't like the toxic climate.

I decide on my own in a way like: read a ton of similar questions and their answers (if they have some) and rtfm to fix the problem. Or simply ask my friends.

If you must you can imply that SO is not ones friend. At least for me this holds.

> There are so many nuance to a problem that the approach of "one question - one answer" is simply naïve.

Yet I have marked hundreds of questions as duplicate which was often welcomed by the person who asked the question, so clearly it's not fundamentally broken, even though there is some friction at times.

I've had two questions closed as duplicates--both involved situations where how I asked the question didn't remotely look like the duplicate question (no way I would have found it by searching) but it actually had the same answer.
> Posting duplicate questions hurts the community. The whole idea is to have one answer for one question. If the answer changes over time it is updated. This does not work if there are 500 identical questions.

And what about when the nuance of the question changes, supporting points are different because tech has changed, so the original question is no longer accurate even though it's the same problem?

If you want to have this discussion you're going to have to provide a concrete example.
IMO it's kind of strange to want evidence for something that can very clearly occur. Do you really deny that it happens? As it is though I don't have any examples handy, maybe someone else will be able to provide one for you.
If you show an example i can either try to explain why it is the way it is or i can agree with you that it is unreasonable.

I can not say anything useful about a hypothetical scenario you make up. Maybe it happens, maybe it doesn't, we don't know.

> Nah doesn't work. I've had the same thing. I've even linked the similar question in the original, saying how I needed more. Editing the post doesn't bring it back to life. It stays dead. Like pulling the knife out of a stabbed corpse. They are still dead.

The only way I've found around this is to wait 10 hours or so and repost, hoping the original people who had an issue are now sleeping.

> Back in the day, questions were all valid. It was the way devs should work. If you posted an identical question, so what? Who did it hurt?

That was never the case; duplicates were around back in the day as well.

The idea is to build a long-term repository of useful questions, and millions of duplicate questions if not helpful for that. If I answered question X with great detail and lots of context then I really don't want to do that every month, so, you mark it as a duplicate, and in principle everyone wins with this.

I'm not saying the process always works brilliantly in practice and I agree there are too many curmudgeons around and that sometimes questions get closed as duplicates too fast, but the basic idea of it is fine.

This edit feature might be a privilege of some sorts. I have a decent amount but when they closed my question as a duplicate I was not able to edit the question.

I don't know why they are they doing it, clearly posting a reference to another question entirely. Why not just let the question hang, it feels almost bot-like behaviour.

I mean, maybe the original post was an emotional exaggeration, but the systems designed like Stack Overflow where the vendor out sources their day-to-day operations to the member of the public (free-to-join for all) are prone to this kind of vulnerabilities.

The main problem is, without strong oversight, interest group starts to develop and then you got a Cambrian Explosion of motives. And as result of that, fragmentation of experience starts to happen (and maybe even corruption???).

Maybe a centralized management style is better for some of these services. At least so the users can directly talk to someone who's contractually responsible to serve them, instead of to another user from the public who is privileged but unpaid.

Is this the first time that the often criticized SE company is holding up to the users, whilst the moderators went on to strike the company and its users?

As I see it, the mods want to continue on their ChatGPT removals and bans, the company is against that and wants a lighter approach.

def int2binary(num): lis=[] while True: if num==0 or num==1: lis.append(num) break lis.append(num%2) num=num//2 lis.reverse() num='' for i in lis: num+str(i) return num print('Number\t\tBinary Form') for i in range (100,1001): print(i,int2binary(i),sep='\t\t')