-9

I am new to the Stack Exchange community, and since the day one, I struggled to understand one of the most fundamental aspects of its mission. Many posts on Meta have been written to broadly discuss this topic. Yet, an answer to one of the most important questions is not clearly delivered to every new user. On the contrary, trying to find the answer leads to some contradictory and confusing results.

To the point, my question is:

Is Stack Exchange a database of knowledge (in the form of questions and answers) or a help site where expert volunteers contribute their time to help others with difficult questions?

Now, despite the fact that there is likely to be a gray zone between those two missions, everyone should be clear as to which one is more important before they press the "Ask Question" button or decide to criticize a question.

Differences

I think that most of you already realize what the consequences of each mindset are, but let me summarize:

###1) SE as a Database of Knowledge

Principle:

  • If a question, no matter how simple, has not been asked yet, it should be asked.

Some arguments:

  • Each of us needs a quick reminder of simple solutions from time to time and SE makes it much easier to find a solution than to browse through manuals. Isn't it first in Google search results for such questions anyways? How much time did such questions save each one of us at some point in the past?
  • If none of us are experts in everything and every one of us (even secretly) actually benefits from simple questions, why shouldn't we contribute answers to someone else's simple questions and tolerate those as well?

###2) SE as an Expert Help Site

Principle:

  • Questions should be asked only about problems for which the answer cannot be found using other means, i.e. extensive research online, reading manuals etc.

Some arguments:

  • Realize that by asking a question, you waste someone else's time.
  • The person answering your question is doing the work for you (for free), so ask only as the last resort.
  • The flood of simple questions decreases the SNR and makes it more difficult to find those intriguing, brilliant questions and answers.

I assume that in both cases questions should be: on topic, clear and well written.

Why am I confused?

Opinion of the Community

The community and the FAQs seems to be giving mixed signals. First, look at https://stackoverflow.com/tour, which clearly states:

With your help, we're working together to build a library of detailed answers to every question about programming.

Every question? Well, that sounds pretty much like 1) to me. But what does the community really thinks? Just check out the most voted answer to the question How much research effort is expected of Stack Overflow users? (revision 1):

A lot. An absurd amount. More than you think you are capable of. In fact, asking a question on Stack Overflow is the absolute last thing you ever want to do. You want to avoid it at all costs. You want to think of it as a horrible shame that will forever haunt you and pass down from you to your descendants.

That sounds very much like 2) to me.

Actual Users Behavior

Let's look at some selected, most voted questions on SO (all time):

and maybe a more recent example of this month:

According to 2), those questions should make the OPs and their descendants live in horrible shame forever. Yet, those are some of the most voted questions. Votes mean reputation. And reputation is here to promote and reward certain behaviors on the site, isn't it?

So which is it?

Or maybe there is a discrepancy between what the founders of the site wanted and what the community believes is right?

And please do not say it's somewhere in the middle. New users need simple guidelines, call them ideals they should strive to follow. Then, reality comes and things tend to fuzzify naturally anyways. But is there a consensus about which of those should be emphasized more?

EDIT:

So, the question attracted a reasonable amount of viewers and provoked a very useful discussion (to me at least). However, I am also trying to understand the reason for the downvotes, which I believe on Meta mean that I have made claims that the community disagrees with. Therefore, I feel that I need to clarify my position a bit:

  • I am not a proponent of any of the two views (in this question at least), rather, I have noticed a contradiction which I believe exists and has been confirmed by others in the discussion.

  • I perfectly understand that the world is not black and white, but I deliberately tried to be more strict about taking a side here. My observation was that: whenever someone openly tried to raise a similar question, the answer was: it's in the middle or it just doesn't work this way. That would be fine, however, many other questions that hint on the problem (some of them already referred to in this discussion) show that people have very (VERY) strong opinions about this problem. I can also see that in the comments given to questions asked on SO and the "close" votes. Now, acknowledging there is no consensus in the community is something different than saying that there is a consensus that it's "in the middle".

  • I really hoped that a variety of answers to my question will make new users who care and Google that issue either:

    a) understand that both types or questions are acceptable and both are likely to be criticized and for instance feel better whenever they are referred to as rep-whores for answering a Googlable question or when they are bashed for asking one while thinking they are contributing to the body of SO knowledge.

    b) understand what is expected of them in case there actually is a consensus.

28
  • 12
    Not my downvote, but all your examples are six to seven years old. They are hardly suitable to discuss today's guidelines. Commented Jun 6, 2015 at 18:08
  • So the answer is, things have changed? To what then? What are the current guidelines? Or are you saying that all those votes are old and the question of what big o is used to be more difficult to answer in the past? No matter how old they are, they still serve as examples for new users. Commented Jun 6, 2015 at 18:18
  • Besides, those examples are not really the point, since they are clearly justified by the pretty current statement on stackoverflow.com/tour Commented Jun 6, 2015 at 18:22
  • 6
    It's somewhere in the middle. :) Both are correct to some extent. The ideal goal is to build an archive of great questions and answers that everyone can search, and ideally never needs to ask a question again that exists there. But the place is also a help site and full of duplicates because it's human beings asking and answering the questions and helping each other, getting things wrong, and being lazy and/or incompetent. There is no clear consensus on whether totally easily Googleable stuff should be a SO question, or downvoted to oblivion; that's probably why no clear guidelines exist. Commented Jun 6, 2015 at 18:23
  • 8
    Don't stop reading at "every question about programming". Some of the subsequent requirements will exclude a large segment of questions. At best read it as "every question about programming, following the guidelines below". Commented Jun 6, 2015 at 18:27
  • Ok, in that case, why is my question being downvoted if it touches a problem that does not have a clear solution or for which clear guidelines exist? :-) Commented Jun 6, 2015 at 18:28
  • What will happen if the answer to your question is 1, 2 or in the middle? Are we going to have a follow up question to instruct the community to follow the rules implied by 1, 2 or the middle? Commented Jun 6, 2015 at 18:29
  • 1
    Good luck coming up with those. Commented Jun 6, 2015 at 18:30
  • 6
    Then stick around and watch. You are never ever going to get crystal clear guidelines. Such guidelines are first of all impossible to formulate. And second of all, the things that are crystal clear to some, are muddy, unclear, and open to misinterpretation for others. What you see now is the outcome of how the site has evolved over time. And not only that, it's how the text reflecting this evolution has evolved over time. Follow along on Meta and you'll see discussions every now and then, trying to make even more clear what we thought was already perfectly clear. Commented Jun 6, 2015 at 18:34
  • 1
    @James, I agree. Commented Jun 6, 2015 at 18:38
  • 1
    Especially that it is hard not to notice that there seem to be many very strict rules and guidelines regarding other aspects of SO, while no current voice exists on such fundamental question in a single place, even if that voice is "there is no existing consensus, meaning, both behaviors are acceptable and both will also be criticized" Commented Jun 6, 2015 at 18:42
  • 4
    there is no existing consensus, meaning, both behaviors are acceptable and both will also be criticized Ha, that's nicely put. I'd say that is a fairly accurate characterization of the situation. Commented Jun 6, 2015 at 18:43
  • 2
    Oh yeah, votes are pretty much all we have as far as measuring "goodness" here; I accept that, and they work pretty darn well. I'm just saying that it's not a perfect signal. Commented Jun 6, 2015 at 19:29
  • 2
    Oh, also: "a discrepancy between what the founders of the site wanted and what the community [wants]". Never mind the users, there was discrepancy between what the two founders wanted! Commented Jun 6, 2015 at 19:31
  • 1
    Good question. Interesting to know that people discussed these topics already that early. The exact combination of both goals and how they interact is indeed at the heart of what makes a good Q&A site, a great site or a failing site. We were quite good, but ultimately hit the ground quite hard in the last years. More discussions back then may have helped, but then nothing beats experience. With hindsight some of these questions should be answerable. Commented Apr 1 at 17:11

3 Answers 3

Reset to default
12

The two purposes for the site as you define them aren't mutually exclusive. The first point is indeed the primary goal. A repository of knowledge is what we're trying to create. But that does by no means imply that all conceivable questions are on topic. And they don't need to be.

There is no real value in hosting the answer to every conceivable programming question you can think of. Not only that, you would end up putting the bar so low, that you would end up drowning the few gems in a sea of mediocrity.

So we try to put up a set of guidelines, trying to optimize for pearls. For those things that aren't trivially found elsewhere. We're not a replacement for basic teaching material. But we're there if you're stuck on an actual programming problem. Something you've been scratching your head over for a while. That has got you stumped. And not the "I googled for 5 minutes" kind of deal.

Crystal clear guidelines aren't going to happen.

Well meaning users, moderators and employees alike have spent countless hours debating the scope of sites on Meta. They have come up with boundaries and texts describing them. The FAQ is an ever changing entity and the Help Center has seen countless revisions. So now everything is perfectly clear, right?

No. And it won't ever be. The world isn't black and white. And even if something seems to be crystal clear to you, others will interpret it differently. And not because they are wilfully ignorant. It's just that when writing such content, perfect clarity doesn't exist. All you can hope is that you've placed some fence posts in the right locations, and that the chains in between don't allow for too much movement.

And to complicate matters even further, even in the hypothetical case that there's no room for interpretation, perfect enforcement is even more unlikely to happen. You care. We care. But the far majority of visitors to the site doesn't know, or simply doesn't care.

And I guess that brings us to some of your example questions and others like it. Although their votes are the results of several years of vote accumulation, there are others like it that seem at odds with what the site says it's about, but which have still received their significant amount of upvotes. Why? The site has strong Google-fu, and people found it valuable. They appreciated it. And whether or not the post was in line with what the site would like is often of secondary importance at best.

5

It just isn't a simple formula that you can stick in a bottle and hand out to anybody who wants to drink. There have been plenty of cases of ho-hum questions turning into fantastic Q+A that's been a major resource to many programmers. Not infrequently after having the question edited and reformulated several times. And fantastically detailed and researched questions can easily never get an answer and left forgotten and ignored.

The law of averages however applies strongly. The odds that a poor question ever turns into helpful Q+A are just not very good. A starting point for good Q+A will always be a decently researched question with strong participation from the questioner, responding to comments and tweaking the question, and plentiful help from multiple SO users, editing and answering.

Do avoid staring at ancient Q+A to get insight. Content at SO had a very high ranking at Google and the ones you quoted got massive number of views. 701798 for that Python question, that doesn't happen very often. Not anymore. Given that three quarter of a million programmers looked at it and only 0.49% of them thought the question was worthy of a helpful vote does put a better perspective on it. Not that impressive.

Q+A like this one would be more representative, 13 votes out of 230 views, ten times better than the Python question. It does slack off over time however.

-13

There is a real contradiction here, and it’s not subtle—it’s structural.

Stack Overflow still presents itself as:

“a library of detailed answers to every question about programming”

and historically:

“build good answers to every imaginable programming question” and "as frictionless and painless to use as we could make it”

That describes a system where questions are encouraged, because they are the raw material from which the knowledge base is built.

But in practice, the message many users receive today is the exact opposite:

  • don’t ask unless you’ve exhausted all other options
  • don’t ask anything “simple”
  • don’t ask anything that might already exist in some form

At that point, the model quietly shifts from “ask and build knowledge” to “avoid asking unless you meet a high and often subjective bar”.

And this is where the contradiction becomes impossible to ignore.

A large portion of the most valuable content on Stack Overflow consists of questions that are:

  • simple in hindsight
  • widely applicable
  • easily searchable after they were asked

Under the stricter interpretation that is often enforced, many of those questions would likely never have been asked—and the “library” would be significantly poorer for it.

So this is not just “somewhere in the middle”.

It is two competing models:

  • A system that needs questions to exist in order to create knowledge
  • A system that increasingly treats questions as something to limit and filter aggressively

Those two directions do not naturally align.

What makes this more notable is that even the leadership of Stack Overflow has recently acknowledged issues around participation, usefulness, and direction - see: I’m Jody, the Chief Product and Technology Officer at Stack Overflow. Let’s talk about the site redesign

That discussion explicitly raises concerns about:

  • declining participation
  • whether the current model is still serving users effectively

Which strongly suggests that this tension is not just perceived—it is recognized at the highest level.

So the confusion new users experience is entirely understandable.

They are told: “help build a library of answers to every question”

but experience:

“don’t ask unless your question survives strict and sometimes inconsistent filtering”

At that point, the problem is not that users misunderstand the system.

It’s that the system is sending two different signals at the same time.

If you look at how the site was originally presented, the main goal was to provide answers to questions. Even today, the tour still says:

“Ask questions, get answers, no distractions.” https://stackoverflow.com/tour

The idea was that people ask questions, others answer them, and over time this builds a useful knowledge base. The focus was on helping people first.

Early descriptions of Stack Overflow also emphasize this direction. For example, the 2008 “About” page states:

“With your help, we can build good answers to every imaginable programming question together.”> “Stack Overflow is as frictionless and painless to use as we could make it.”

https://web.archive.org/web/20081204052725/http://stackoverflow.com/about

This suggests that the original goal was to reduce friction and encourage participation, so that answering questions would naturally build a useful knowledge base over time.

However, over time the community developed a different set of expectations. Questions are expected to be a last resort, simple questions are discouraged, and there is a strong focus on filtering what content survives.

This is reflected in posts like:

How much research effort is expected of Stack Overflow users?

This situation has a real impact . The mismatch between expectations, communication, and actual enforcement may not be trivial and has been noted repeatedly (e.g. Meta discussions explicitly referring to a “mismatch of expectations” and proposals to adjust how new users understand the system). Sample posts show that a significant portion of new users encounter friction such as closures or lack of feedback. Taken together, that may show that the gap may have measurable effects on participation.

For example:

10
  • 6
    FWIW, the community has asked time and time again to adjust the immediate public facing advertisement blurbs to match the rules and even slightly-extensive clickbaits. Even the tour already says "Don't ask about... Questions you haven't tried to find an answer for (show your work!)", so it's not like much digging is needed to realise the kind of six-word summaries aren't meant to be interpreted the way you do here. Commented Apr 2 at 8:58
  • 3
    Also, three out of four of the quotes you show here clearly put answers before questions. Yes, we need questions to answer but ultimately it's the answers that add value. Commented Apr 2 at 9:01
  • @MisterMiyagi I think this highlights the core issue rather than resolving it. If the public-facing messaging bothers the community - and the corp. refuses to change it - that is proof of a contradiction/conflict. If the message is consistently interpreted in a way the community considers “incorrect”, then that’s not just a user misunderstanding—it suggests the messaging itself is not effectively aligned with how the community actually operates. If the biggest font slogan/description regularly leads to mismatched expectations, then the communication is at least part of the problem. Commented Apr 2 at 9:45
  • @MisterMiyagi On the second point, I agree that answers are ultimately where the value lies. But answers do not exist independently—they are a function of the questions that are allowed to be asked. Many of the most useful and widely referenced answers originate from questions that are “simple” in hindsight or that overlap with existing material in some form. If those questions are filtered out more aggressively today, then the system may also be limiting the production of future high-value answers. Commented Apr 2 at 9:47
  • Summarizing point 1: If messaging consistently creates expectations that are later rejected by enforcement, then even if intentional, that gap has real consequences for user experience and participation. Commented Apr 2 at 9:53
  • 1
    I am not claiming that anything is resolved by what I indicated in my first comment. Rather the opposite: the community has repeatedly called out that this hasn’t been resolved. Commented Apr 2 at 10:29
  • @MisterMiyagi It sounds like we’re pointing to the same underlying issue — that there’s still an unresolved gap between how the site presents itself publicly and how it operates in practice. 🙏 Commented Apr 2 at 10:40
  • 2
    As for your second comment-point: I am not aware that the rules have changed to forbid "simple" questions (unless this word hides some other meaning). In fact, a previous close reason that went in this direction ("lacks minimal understanding") was removed years ago. So I am not sure what model you are referring to on that point. What the SO model does include is some filtering of allowed questions by design, and I don’t see how this general feature is debatable – we can’t cover literally every question. Commented Apr 2 at 11:05
  • That’s fair — I agree that there isn’t an explicit rule forbidding “simple” questions, and that some level of filtering is necessary. A point is more about how the system is experienced in practice. Even without an explicit rule, questions that are perceived as too simple, insufficiently researched, or potentially duplicative are often closed. That’s where the tension comes in — between what is encouraged vs/or sustained in practice. Commented Apr 2 at 11:59
  • 2
    I don't think questions are closed for beeing too simple often. Simple questions are often duplicates, too broad, poorly written, or have other issues for which they get closed though Commented Apr 4 at 13:24

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.