[legitster]

God bless the people on Stackexchange who actually answer the question that is asked instead of scolding you for your idea being bad or being pedantic about the way you phrased your question.

Yes, I want to use jQuery. No, I may not be in an active war zone.

[Cherian]

My life was built on people who helped me out like this.

Tangentially, the real MVP is the home depot guy who helps you find the one right-sized screw that costs $0.5….

[ethbr0]

Interesting story on that, Robert Nardelli (HD CEO 2000-2007) fired all the ex-tradespeople that Home Depot employed in their stores, because they cost more than younger and less experienced labor.

Straight from the GE "How to mortgage a company's future for a small boost in the present" book.

There were also some hilarious anecdotes told about him refusing to get out of his car in the corporate parking lot until security met him and escorted him in, presumably because he understood how much employees disliked him.

https://en.m.wikipedia.org/wiki/Robert_Nardelli#The_Home_Dep...

[legitster]

There's no better reminder that only a thimbleful of useful human knowledge is actually found on the internet.

I recently spent 4 hours online trying to solve a carpentry problem and not even knowing what words to use. I finally called my dad and in less than 2 minutes it was solved.

[te0006]

Now _that_ would be a good, hard challenge for ChatGPT and its ilk. Could you post question and correct answer here? Preferrably also something approximating the original version (when you still didn't know the correct technical terms).

[contingencies]

The communication of solutions to untrained audiences through the employment of simplified semantics is definitely an interesting field of linguistics. Visual smacks of Ikea.

[throwawaaarrgh]

it's impossible to squeeze experience into text. that's one thing AI will never be able to replace: 20 years of carpentry projects.

I can show you how to do the thing with the chisel in this particular type of wood and humidity and grain and angle and pressure, but I can't write it down.

[contingencies]

Fair point. In all seriousness though, it does often feel like Youtube is filling that niche quite nicely. It is almost as if the formerly wondrous power of written language to communicate across time and space is being debased and usurped by the power of video production and internet distribution...

[msla]

> There's no better reminder that only a thimbleful of useful human knowledge is actually found on the internet.

Or in books, for that matter.

[m463]

I remember the (reasonably-priced) hardware store back home being full of these guys, and home depot with no help, selling plastic plumbing supplies.

I guess home depot has won, and the employees have enough faq experience to help now.

Now if home depot sold 80/20 supplies...

[dghughes]

I think the majority of what I find useful on Stackexcahnge sites end up being a post locked down but someone manged to sneak in an answer.

[kelnos]

I wouldn't say majority, but I too find a lot of useful answers on questions that have long-ago been locked down as "off topic" or "too vague" or "too opinionated" or whatever.

[tetha]

You also see this in system administration circles/SRE quite a bit. And I mean, that's one of the important skills you have to learn in the operative trade: You should push for clean solutions. You shouldn't use clothesline to transfer power. You shouldn't use an EOL OS to run systems.

But as much as you push, sometimes you need to put on the rubber boots and gloves going up to your shoulders and figure out a somewhat safe way to run some Windows XP based machine controller in an environment. Or wrangle some Java 1.6 thing back into function. Or figure out the least security reduction to support some old system not supporting modern crypto.

And yeah, usually the idea is to put trusted isolation layers around the dumb idea we have to deal with, as the water hose suggested in the article.

[hot_gril]

Ugh. I asked if you could use consumer-grade SSDs with an HP ProLiant equipped with hardware RAID on one of the StackExchange sites, forget which. The only response I got was, "Oh, too cheap to buy the official ones? This website is for professional sysadmins, it tells you that when you sign up." Which btw it doesn't say anywhere.

[elijahbenizzy]

"I'm in an active warzone, how do I add two numbers" "Well, there's this library called jquery..."

[shagie]

There's a substantial difference between the amount of time and guidance an individual can spend with a single question when you're on a site that gets 3.9k questions per day and one that gets 21 (stats from https://stackexchange.com/sites#questionsperday )

[interroboink]

Your phrasing makes it seem like the people answering have the job of "keep up with the rate of incoming of questions." So, if there are a lot of incoming questions, they must reduce quality of feedback, since they are spread thin.

Personally, when I answer a question it's because I want to, and feel I can be helpful. I have no skin in the game with regard to the site's overall ability to keep up with incoming questions. So, I take as long as I need, and do as much hand-holding as I feel is appropriate, not governed by external pressure.

But I suppose there are professional moderators and such who really do have that external pressure, and thus have incentive to give curt feedback, or even to drive people away — thus reducing that pressure, making their lives easier.

As a SO user from the early days, I do miss that feeling of mostly interacting with people doing it "for the love of it," rather than governed by efficiency.

[shagie]

The ability for you to find a good question to answer is in part based on the work that other people do in down voting and closing questions. If you go to the triage queue - those are questions that are being prevented from showing up.

If you go through "newest questions" there are often questions there that can't be reasonably answered without more work to figure out what the problem is.

As I write this, there's a question https://stackoverflow.com/questions/76123197/is-it-possible-... which is apparently a Java and Kotlin question

> Is it possible to achieve Color gradient overlay like Resso app

> It's really going to be interesting (big inline image)

And... is that worth clicking answering in that form? How much time should you spend trying to make it a better question that someone else can answer? Or just down vote it and move on? Would you ask "how does this related to Java?" or "could you explain a bit more about what you've done so far and what problems you've encountered?" - and is that considered rude?

The corresponding part of it is that people who have their question down voted without any information may find it rude. Or if someone suggests a change to the question... they may find that rude too.

And some people find not getting a response at all to their question on a site that is billed as the place to get your questions answered rather frustrating.

In order to make it easier for people who want to answer questions to find questions to answer a lot of questions don't show up. Poke at the triage review queue and consider the additional difficulty of finding a question to answer if those were also present in browsing.

Note also that there are no professional moderators on SO. Everyone there is a volunteer... and thus they're burning out a bit too. While it may be easy to say "well, then they should take a break" - they do... and more questions of questionable quality show up in the feed.

The best way to find good questions to answer is to look at recently asked the up voted questions (and avoid the down voted ones)... but down voting is considered to be rude.

And if you want to help a question by asking a clarifying comment in there for this one that might be interesting... and that one... and that one... and do it for ten questions or so you've spent half an hour... and those comments trying to get some information about how you should answer are seen as rude. How much time do you want to spend asking clarifying questions in comments?

Ultimately, SO is suffering from issues of scale without the corresponding tooling to enable people who are trying to answer to find interesting questions more easily. That's not an issue on smaller sites where you can read all of a day's questions over lunch.

[interroboink]

> Ultimately, SO is suffering from issues of scale without the corresponding tooling to enable people who are trying to answer to find interesting questions more easily.

Yes, that seems like a reasonable take.

I suppose in my imagined ideal reality, people simply don't answer questions that are not asked well, or that don't interest them for any other reason, rather than actively body-slamming those questions.

If this results in a glut of low-effort questions, then the site suffers. As a result, the site has an incentive to provide better tools.

Right now, volunteers heroically stem the flood of poor questions by burning themselves out and sometimes getting bitter. The site still suffers, but in a different, more pernicious way.

I looked at a triage queue question just now, and it was indeed poorly-written. I selected "Needs author edit", and clicked "Submit". Then, I got a pop-up asking "Why should this question be closed?" and I was confused. I don't want to close the question. I don't want to send that signal to the question writer. I want them to improve their question, that's all. I canceled the interaction. So again: agreed about bad tools. Personally, I choose not to use them.

[shagie]

The difficulty that SO has had trouble with is the "then the site suffers." How do you measure that? They want to run some A/B test that allows the corresponding measurement to show that things are better with a change.

However, it feels that the only way that they've really accepted measuring it from a sales / marketing view (as that's what brings in the revenue) is the "engagement" metric. People signing up, asking questions, and accessing the site.

Better moderation tools which would result in fewer but higher quality questions on the site shows up in that measurement of engagement as "worse".

---

The part that you encountered is that "if the question isn't answerable, it should be closed." That in turn feeds other parts of the system. Users are more likely to update their questions if they are closed rather than if they're left open. Other people who answer questions (but rarely engage in fixing up questions) are less likely to click on questions that are closed. People that routinely ask poor questions that get closed start getting automated warnings about their question quality before they ask a question and end up with a question ban if that behavior persists. A closed question without answers or edits to improve it get automatically deleted after 30 days.

Without going in and commenting on a question and then spending time with the person ("why don't you just answer it if you think you know the answer rather than commenting? If you don't like it just don't read it." is something I've seen many times) closing the question is a way to suggest improvements to the question without exposing yourself to users who not infrequently then pursue a... negative engagement with the person trying to help them ask a better question.

I can see about digging more (it's been a long time since I went looking for it) but somewhere on one of the meta sites was a post about the different interactions and the "engagement" metric for new users asking a first question.

The best way to not have them ask a second question is to completely ignore their question - no votes, no comments, no answers. Closing a question results in more people asking a second question that is positively received than having no interaction.

(late edit - did the digging - https://meta.stackexchange.com/questions/216683/what-happens... )

[interroboink]

I may be going in circles here, but when you say:

> The best way to not have them ask a second question is to completely ignore their question...

Leaving aside the issues of interpreting that data[1], and taking your conclusion at face value:

I get the sense that you imply this is a bad thing. But is it?

I agree that if the ultimate goal is "boost engagement metrics," then it's a bad thing. I suppose I just don't agree with that being the ultimate goal. And I sure wouldn't mind if other people in the community de-prioritized that goal, too. My opinion here applies to much about the modern internet landscape, to be fair (:

[1] eg: was their choice to ask a second question caused by a particular interaction? Or maybe users that ask second questions are more likely to ask a good first question, or other explanations and confounding factors? By another reading of it, we could say users who got their first question closed were 2x as likely to leave permanently than those who had no interaction (and this applies to over 2x as many users, so is even bigger in absolute terms). It's rather muddy.

[s1artibartfast]

Thats a reason to not engage if you dont have time.

It is not a reason to sit down and dismiss away questions or type out why you wont tell them.

[shagie]

Today, most people who are doing curation of Stack Overflow are not engaging with a question at all. They down vote and move on as any attempt to help is seen as rude... and just down voting and moving on is also seen by some as rude... and not doing anything and having a question get ignored is also seen as rude or frustrating.

As there are many fewer people answering questions compared to the rate of questions being asked the overall "is stack overflow rude or not" is a "yes." But engagement numbers are up as people keep asking them.

[armchairhacker]

Stack Exchange itself is underrated, it's just that Stack Overflow is really bad.

[shagie]

The moderation tools and expected community involvement in using them was built for a much smaller site with a more active user base... and it works reasonably.

When you then add on top of it "engagement" metrics, scale up the number of questions per day by orders of magnitude without the corresponding scaling up of the community involvement and (to an extent) try to remove the ability for the community to moderate and curate the content then the tools that are left to the community are the social ones (as they can be used beyond the limited number of down, close, and delete votes that one has in a day).

And then you're left with "the way to handle questions where the person didn't even put the title into Google to search first is to be rude to them." It's not a good thing, but without the barriers to entry being implemented in code they are erected by reputation and social forces instead.

It isn't a good thing - and it would probably be much better if those barriers were put in place through some other means... but as long as engagement is the measured metric and ad impressions are the income, having company's developers implement it is a non-starter and you're left with the community using rudeness as the moderation tool of last resort.

From A Group Is Its Own Worst Enemy ( https://news.ycombinator.com/item?id=27706413 )

> Four Things to Design For

> ...

> 3.) Three, you need some barriers to participation, however small. This is one of the things that killed Usenet, because there was almost no barrier to posting, leading to both generic system failures like spam, and also specific failures, like constant misogynist attacks in any group related to feminism, or racist attacks in any group related to African-Americans. You have to have some cost to either join or participate, if not at the lowest level, then at higher levels. There needs to be some kind of segmentation of capabilities.

> ...

> 4.) Finally, you have to find a way to spare the group from scale. Scale alone kills conversations, because conversations require dense two-way conversations. In conversational contexts, Metcalfe’s Law— the number of connections grows with the square of the number of nodes—is a drag. Since the number of potential two-way conversations in a group grows so much faster than the size of the group itself, the density of conversation falls off very fast as the system scales up even a little bit. You have to have some way to let users hang onto the “less is more” pattern, in order to keep associated with one another.

[eitland]

I have said it before I think, that I am sure Stack Overflow absolutely can, if they want to, reduce the amount of low quality questions they get.

The problems Stack Overflow has seems to me to be very much self inflicted, caused by the decision to optimize for political games instead of optimizing for solving problems.

[shagie]

"If they want to" runs into issues that as a company, they're measuring their success by engagement.

Adding that barrier to participation would in turn drive down engagement and advertisement impressions.

From the corporate standpoint, doing that (or anything like it) translates into a loss of revenue.

[eitland]

Maybe. Personally I think software engineers are a much more valuable audience than college kids.

Also I think a lot of what happened was rampant deletionism, that I personally can't see any good reason for with todays storage prices.

[shagie]

The issue isn't the storage... but rather the difficulty of using a search to find a good question that has been answered.

Do we need 10,000 questions about how to handle a NullPointerException in Java? https://stackoverflow.com/questions/linked/218384?lq=1

If not, then it is probably appropriate to delete more of them and that may be seen as rude by people who have asked a question about a null pointer exception which gets deleted.

We complain about how Google has gotten worse with search because its harder to find the content that we're after... and at the same time say that we want to keep all that content around on Stack Overflow which in turn makes it harder for people to use it as a "this is where you look to find an already answered question."

[eitland]

Back when the answers were still there and Google still worked I managed to find them just fine.

Many others clearly did too as many of the questions they removed or tried to remove were massively upvoted.