Live data from Hacker News

Stack Overflow is almost dead

blog.pragmaticengineer.com

131–140 of 285 posts

Re: Stack Overflow is almost dead

#131

Earlier quoted context omitted.

The presumption is that things will improve over time, but the big difference in my experience is the assistance I got from SO _worked_ the vast majority of the time, whereas various LLMs I have used generate unusable, misleading, or unreliable results pretty regularly, increasing as complexity or rarity arises. As human-driven knowledge bases backed by actual experience are replaced by inference from models who rely…

Personally, I am getting extremely tired of ChatGPT hallucinating npm packages that don't exist, or package imports that do not exist

Get a good service. Running it on your old gaming machine?

Re: Stack Overflow is almost dead

#132
post #56
post #9

I find it interesting that the current StackOverflow moderators tend to say "in the past we used to accept too many questions but it was never the goal, so now we are doing it as it was meant to be". Sure, but in the past, StackOverflow was growing, and now it's dying. Maybe something was better before, when "it was not done correctly"?

I think these sorts of things are just an unfortunate side effect of scaling. The bigger you get the more people get lost in the bureaucracy. However if you don't build up the bureaucracy the system collapses under its own popularity. Wikipedia has a similar issue where editing declined around 2007, which is often blamed on stricter enforcement of rules, more complex rules, etc. I think its just a natural stage of gr…

The "good" thing is, they're back to 2009 levels of postings. Now obviously that's what the mods let through but my guess is that traffic to the site is down precipitously as well. They can roll back their bureaucracy and head back to a lean path that worked for them in the past.

But I don't really think that's the problem. Reading zahlman's responses in this thread makes me think that the mods fell into the age old trap that's happened since Usenet, IRC, and still happens to this day wherever there's mods: they got tired of doing unpaid labor and instead of deciding to quit decided to become meaner and stricter. The age old mod trip.

Re: Stack Overflow is almost dead

#133

> 2014: questions started to decline, which was also when Stack Overflow significantly improved moderator efficiency. From then, questions were closed faster, many more were closed, and “low quality” questions were removed more efficiently. This tallies with my memory of feeling that site moderators had gone on a power trip by closing legitimate questions. I stopped asking questions around this time because the site…

Yes I felt the same around then. It seemed like stack overflow was sending the message to not ask questions anymore. It was really weird.

Same. I've never been a huge StackOverflow user, but it is so irritating to search and find your exact question on StackOverflow, often as the top result, only to see that it was instantly shut down two years ago as a duplicate of some other question in another context with inapplicable and useless answers.

It is frustrating not only because you can't get instant help, but also because it shows the futility of even trying to post on there.

Re: Stack Overflow is almost dead

#134
post #128

Earlier quoted context omitted.

The people who have been around for 10 years or so who disagree with the basics of question closing policy (or who act without any apparent awareness of it) are even worse than the people seeking a quick fix for their problem. Because they flood the site with inherently low-quality answers to low-quality questions. In doing so, they dilute higher quality content (it becomes harder to find with search engines, because…

> The people who have been around for 10 years or so who disagree with the basics of question closing policy [...] Because they flood the site with inherently low-quality answers to low-quality questions I see that you just don't hear the complaints. I don't hear people who have been around for 10 years or so complaining because they can't answer to low-quality questions.

> I don't hear people who have been around for 10 years or so complaining because they can't answer to low-quality questions.

Because they usually can. There aren't enough curators paying attention. Stopping them generally requires three curators to find the same question and act on it before an answer is posted. And the person writing the answer also gets a grace period for in-flight answers.

Then when they finally get inconvenienced, they come to the meta site and make perhaps their first attempt in over 10 years to even find out what the policy is. Often they have a bad experience with this, loudly complaining as if they already know the policy while never having made any attempt to learn it, and being surprised to find out they're wrong. Sometimes they even try to make a meta post on the main site.

Re: Stack Overflow is almost dead

#135

Earlier quoted context omitted.

Personally, I am getting extremely tired of ChatGPT hallucinating npm packages that don't exist, or package imports that do not exist

Get a good service. Running it on your old gaming machine?

Are you implying he is running ChatGPT from his gaming machine?

State of the art models constantly generate bullshit, even if I find them generally useful. Your blind hype does nothing but make people more skeptical of it.

Re: Stack Overflow is almost dead

#136
I have mixed feeling about this.

It has helped me in the past but yet, I could not reply nor post anything back to help others when I knew the solution because of the way how it works.

To make matters worse while working in IT, I worked with a guy that didn't know anything, if there was no SO post about the problem, the guy couldn't fix the problem.

I have been using Perplexity AI and it has been awesome, and it does provide all the sources it used making it easy to cross check the answers. It has helped me to speed my python learning curve, I am not using search engine anymore, and SO has the problems mentioned above so I have zero interest in using it.

Also, the website layout is a mess, I have to use uBlock Origin with a ton of element picker to stop loading half of its crappy.

Re: Stack Overflow is almost dead

#137

My heart goes to the stack overflow community which has always been very kind and helpful, essentially working for free. As a self-taught developer since the age of 8, I literally grew up learning how to code through SO, asking hundreds of questions and answering many more. So many bugs that would take 2-3 days to fix would eventually find their answer through it. But now ChatGPT does that in minutes… so it’s for the…

Are you talking about Stackoverflow? Every time I asked a detailed question it would be closed within minutes. I'm not surprised it's on the way out.

I've asked dozens of questions on SO, and never had a single one closed. I hear your sentiment often, but have no idea whether my experience or yours is more common.

I've had 3 deleted by Community bot as abandoned, but since they were over a year old when that happened, I couldn't care less.

Re: Stack Overflow is almost dead

#138
post #130

Earlier quoted context omitted.

>You keep saying everything is great. I say it's fine, because it is. I say that a reduction in question volume has advantages in terms of accomplishing the site's goals, because it does. There are many things about the site that I'm unhappy with, mainly to do with initiatives the staff are taking that are also very much not true to the site's goals or purpose. > My observation is that I used to be on SO every day, a…

> ... And? And as far as I can see, it is dying. > If the amount of questions went to zero per day I would still not consider this a problem. So on the one hand you find it okay to delete old questions, and on the other hand you find it okay to not add new questions. But it's not dying. > it would become easier to find what you want. It has never happened to me that I could not find what I wanted on SO because there…

> And as far as I can see, it is dying.

And why are you the one who gets to make this judgment?

> So on the one hand you find it okay to delete old questions, and on the other hand you find it okay to not add new questions. But it's not dying.

Yes.

You write this as though you think there is a contradiction here. I genuinely don't understand why. There is no contradiction here.

QuickDraw wasn't "injured" when it lost 2000 lines of code, either (https://www.folklore.org/Negative_2000_Lines_Of_Code.html).

> It has never happened to me that I could not find what I wanted on SO because there were too many similar questions.

Back when I was trying to sort out the mess more actively, it happened to me daily. I distinctly recall multiple instances of spending hours at a time tearing my hair out over it, and complaining in the corresponding chat about the terrible questions, the unintentional clickbait, and the sensitivity of search engines to minor variations in the query.

> closed by people who obviously had no understanding of my question (together with its answer).

This is said by perhaps 90% of people complaining about their question being closed, and trivially shown to be incorrect in perhaps 90% of those cases.

But also, "people can't figure out what you're trying to ask" counts against your question. By design. Because questions are expected to communicate clearly. So that other people who read them don't have to waste their own time making sure they're in the right place.

Of course, there are other reasons a question might not be understood. But it's not hard to distinguish between "this person can barely write coherent English" and "I don't know anything about this technology". People are, broadly speaking, just not going around the tags for technologies they don't know about in order to close questions. What on Earth would they get out of that?

> What I am saying is that SO has become a place where even good questions get closed. By people who know better, like you.

Again: please show a link to an example of a question that you believe was unjustifiably closed, and make sure that you can clearly explain, in terms of existing policy why you believe the closure was invalid.

Do this on https://meta.stackoverflow.com, where it belongs.

Or if you have done this in the past, link me an example of that. That's fine here.

Re: Stack Overflow is almost dead

#139
I used to answer questions a lot, by around 2013 I had answered maybe ~12% of all HTML canvas questions ever asked. To me it declined a lot sooner, 2014 really does feel like the right inflection point.

There was a belief, sometimes unstated but often explicit, that no more (serious) discussion is really to be had, and further wondering how can one stop people from asking. It became difficult to discuss anything if there was even something vaguely related asked before. It was not possible to discuss something you knew the answer to, but did not know why, or wanted to hear arguments for which of 5 ways might be best. All (to me) very worthwhile technical discussions. Totally shut down.

Re: Stack Overflow is almost dead

#140
post #130

Earlier quoted context omitted.

> ... And? And as far as I can see, it is dying. > If the amount of questions went to zero per day I would still not consider this a problem. So on the one hand you find it okay to delete old questions, and on the other hand you find it okay to not add new questions. But it's not dying. > it would become easier to find what you want. It has never happened to me that I could not find what I wanted on SO because there…

> And as far as I can see, it is dying. And why are you the one who gets to make this judgment? > So on the one hand you find it okay to delete old questions, and on the other hand you find it okay to not add new questions. But it's not dying. Yes. You write this as though you think there is a contradiction here. I genuinely don't understand why. There is no contradiction here. QuickDraw wasn't "injured" when it lost…

And why are you the one who gets to make this judgment?

the reality makes this judgment. something that was worth billions of dollars could probably be bought for $50m (this is too much…). a definition of dead

Post reply on HN