Stack Overflow technology makes me write bad answers

Question

Update: After all those years I came up with the idea, how to make Stack Overflow a friendly and helpful place.

Today I had some time and felt like writing a relatively good answer on Stack Overflow; explaining both the underlying mechanisms and good practices, unlike usual declarative answers in a "do that" style.

It took me like an hour of time, $50 if I were free lance. It'll bring me nothing but 3-5 occasional views, and - if I get lucky - some ignorant comments.

You know, it's quite discouraging.

Instead of pride for the job that was done, I feel frustrated. I know, this answer may offer help to the OP, if he ever notices it (being already satisfied with other answers). So the only fate the answer will meet is to sink forever, below thousands other questions, falling like snow.

So, being a programmer, I am thinking of efficiency. It's a waste of time to spend so much effort to answer one question when there is no way to [actually] reuse it. So, I wouldn't do it again but just turn back into short answers mode without explanations.

Seeing other answers, I don't think I am only one with such feeling.

The current policy is driving knowledgeable and educational users away, yet it attracts inexperienced ones in numbers. Ones who will tell you to use [some technology] which they never have laid their hands upon. Ones who write answers just because they have learned some "good practice" from other questions, and this "knowledge" becomes an endless source of rep points.

The idea of Stack Overflow is rotten. It was excellent when trees were high and traffic was low. It turned bad now; it has become a honeypot for all the "enthusiast programmers" of the world, eager to share their 2 cents faster than you can say the word "close".

You're killing a great resource. Despite all the nice words and proper declarations, it discourages the reuse of the knowledge. And encourages fast on-site answers.

Of course, the rules will tell you contrary. It's all nice and proper (if you don't mind it's strict mutual exclusivity). But mechanically, technologically it works in the way I described.

I tried to write a tag wiki. You know, it was biggest waste of time in my life. As a matter of fact, it is not a wiki at all; you can't even link to a certain section. Not to mention that nobody ever has the idea to read tag wikis.

Reference questions? Don't make me laugh. There is no good mechanism to store and categorize links. Favorites, although intended for the different purpose, are a chunk of crap when you want to use it as a reference. So, you can't compete with 30-seconds answers: a question will get 5 answers before you can even find an appropriate link.

Okay, I can organize links myself but, there are some problems:

1. I am just a human. I cannot remember everything
2. I am just a human and I cannot answer questions at rate of 1 qpm.
3. One person cannot fight the Tide. It's helpless. When technology makes people writing fast [incomplete] answers - no good will of proper declarations can do a thing.
4. Even if a question gets finally closed, it won't hurt anyone - the hydra grew two more already.

You'll call me a brute for insulting honest users and innocent volunteers, but that's fact:
There are much more inexperienced users than experienced ones. And inexperienced users have much more spare time to hang over than experts who have their job to do. So, from that fact we can conclude that most of the answers on Stack Overflow have a low quality. Nobody cares. But I do.

• There should be a technical way to limit the activity of not-so-knowledgeable-yet-eager-to-answer users.
• There should be a technical preference for the closing a question over answering it (have a silly answer for the silly question? welcome to comments).
• There should be a technical solution to accumulate and structure knowledge, a system that lets you choose a good answer instead of writing it by hand.

That's the only way. Otherwise Stack Overflow will remain an automated dispenser of bad practices.

Well, I just felt like letting it out again.
Thank you for reading.

Answer 1

TL; DR

Don't rant about the system. Pick a strategy that works for you, and stick with it. Your goals should be:

1. Make the Internet a better place to find answers to questions.
2. Treat rep as a long-term measure of the value you add. It's a marathon, not a sprint.

If you feel the need to compete, rather than just being a "knowledge philanthropist," then compete in a category that suits you.

Your "Question" Is a Rant

When you boil down all the information in your post, it seems to be mostly a rant against:

1. Fastest gun in the west answers. You have a lot of company on this one; there are many, many related questions and answers on this topic.
2. Competition for votes in highly-populated tags, or eyeballs in low-population ghettos.

Answering your question would be a waste of my time, if I view it as helping just you. However, since I'm interested in helping the broader community, I'm still going to write a longer answer anyway.

Gamification Means "Pick a Strategy"

Stack Overflow is built around the idea of gamification in knowledge-sharing. Like all games, there are rules to the game and ways to abuse the spirit of the rules without cheating.

When you say:

It's a waste of time to spend so much effort to answer one question, when there is no way to [actually] reuse it. So, I wouldn't do it again but just turn back into short answers mode without explanations.

you're inherently complaining that you can't "win" at Stack Overflow using both the fastest-answer strategy and most-canonical answer strategy simultaneously. This is largely self-evident, and so is the solution: pick a strategy that works for you.

Some Strategy Rules for In-Depth Answers

There are many strategies, and variations on each strategy, for participating effectively on Stack Overflow. Creating in-depth or canonical answers is a viable strategy, but you need to be smarter about it.

Whatever strategy you choose should make the experience fun for you and useful to others. If you aren't meeting both those goals, then you've missed the point of the site.

Personally, I'm not interested in playing fastest-gun, so I don't. I answer questions in all the categories that interest me personally, but I certainly take context into account when deciding what to answer in-depth, and how in-depth a given answer should be. I follow some very basic rules about what questions I answer in-depth:

1. They need to be good questions that won't attract 5,013 one-line answers.
2. They need to be questions that don't have big "close me!" targets on their backs.
3. They need to be applicable to a larger audience. Even questions that aren't closed as "Too Localized" may be too narrow for an in-depth answer.
4. They need to be in middle-of-the-road tags. Tags with no followers gain very little rep, and tags with too many followers are usually (but not always) fastest-gun magnets.

That's it, really. And it works for me, both on busy sites like Stack Overflow and on smaller sites like Project Management Stack Exchange. Since you aren't me, your mileage will obviously vary, but I think the core concept is sound.

Generalizing the Advice

I hope this long, in-depth answer helps you personally. If not, I hope it helps someone else with similar concerns. Either way, I'd like for everyone to get the most of the Stack Exchange experience while making the Internet a better place.

Answer 2

I feel your pain.

Following the terms of CodeGnome, there are two opposing factions: the fast-answerers, and the most-canonicalers. You're in the most-canonical camp: you like to write detailed, documented answers, but you find that by the time you've finished typing them, three other people have typed the same two-line answer which obeys the admonition of not just posting code by adding “try this”. And if you do write a long, interesting answer, the question gets closed as a duplicate of a similar question with only quickie answers.

Fortunately, there are a few ways to compensate. First, take the long-term view: most of the quickie answers gather 30 upvotes in 12 hours and 2 more in the next year. Longer answers get their votes in the long run. Second, if you've written a good answer, you'll be able to recycle it elsewhere: copy large swathes or link to it in other Stack Overflow answers; copy it on your blog; use it on other forums; flaunt it as the definitive answer on the subject. Even use it for a practical problem of yours one day — it's a nice feeling when you need an answer quickly, search on Stack Exchange, and find an old answer that you'd forgotten posting and that addresses your current problem in addition to the asker's. Third, occasionally the asker does come back and switch his accepted answer to the longer one.

There is also the intangible benefit (or annoyance?) of seeing comments like “I wish I could vote +200!” or “Thanks a lot for your insight.” or “Thanks for the detailed explanation!”…

On the topic of duplicates, we are now explicitly encouraged to close questions as a duplicate of questions with good answers, and in particular a question can only be considered a duplicate if it has an upvoted answer. This should help move away from the “older question wins” habit (which was never a rule, but is nonetheless blindly followed by some), towards the “best answers win” which makes the site better.

That does leave a lot to be desired. The attitude to quick answers is summed up in the debate on fastest gun in the west (FGITW), where what was for long the highest-scoring answer states “I do NOT want to, in any way, discourage the quick and dirty answer.” (I hope you've downvoted that call to mediocrity.) Please stick around and help us change that culture into one that promotes good answers.

Stack Overflow is not a forum. An answer isn't just for the asker: it's for life.

By continuing to post long answers, you do make a difference. On Stack Overflow, it feels like each answer is a drop in the ocean. But drops is what the ocean are made of. On smaller Stack Exchange sites, I do post long answers, and I think I come out ok on the reputation front (I could do better in rep per time spent, but that's not what I'm here for), and I do encourage others to do the same.

So it's not the technology that makes you write bad answers: it's the culture. Admittedly the technology doesn't help, but it doesn't particularly hinder. And you can change the culture, if you persevere.

Answer 3

There are two things Stack Overflow tries to be:

• A collaboratively edited repository of programming knowledge
• A place to get a quick and dirty answer

For experienced programmers, it may be hard to understand the importance of the latter. You're stuck at your desk Friday afternoon with some trivial issue like a syntax error. You can either:

1. Spend the next half hour researching your problem, hoping to solve it.
2. Post it to Stack Overflow and have someone else solve it in seconds, and go to lunch early.

That's why some people love Stack Overflow. It may be regurgitated knowledge, but it's easier than doing your own research or learning how to apply an answer to someone else's question to your more specific situation.

People like interaction and tailored solutions, not reading some long difficult boring manual. You may call it lazy, but in many cases I can't blame them. You can RTFM all day, but it won't click until someone explains it to you. Then suddenly you have the power to explain it to others and want to share your knowledge to help people (to the best of your limited ability) and have a little fun watching your reputation score go up a few votes.

There is little reward in writing tag wikis; I agree 100% with that. They are practically invisible. Plus, you can't get upvotes (or even acknowledgement) from writing them, so the incentive for rep-driven users (most of the user base) is gone.

From what I've seen, self-answered or proposed "canonical" questions are rarely welcomed; rather people cry for "community wiki" (AKA rep envy), downvote, or close as "not a real question". A while back, someone tried to create a reference for JavaScript operators, similar to the PHP one, and it was downvoted heavily, then closed and reopened several times. Currently, I cannot even find it, so I assume there's little chance anyone else can either.

Over the past week or so, it took several users' effort (including a dedicated chat room and browser plugin) to simply close questions caused by simple typos ("You named your variable i but used j."). We managed to remove nearly 1000 of them. On the one hand it's great that people pitched in. On the other hand, I still feel like it was an enormous waste of time, especially considering all the effort that was put in, only my obsessive-compulsive tendency compels me to keep going.

The close vote queue is practically a joke. With about 50K questions, it's estimated that it might be caught up around October 2015, but I think even that's a dream. I've seen my close votes wither away on so many questions, because the post was so far back in the review queue.

Part of the problem is perhaps the "technology" or design, but the other part lies with the users. Don't get me wrong, I love Stack Overflow. That's why it sometimes makes me sad.

Apologies if this has been an unfocused rant. Personally, I've (just recently) decided to take a rest and not use Stack Overflow for a while. I'm getting burnt out on playing janitor and feel my time is better spent on work and family. I'm sure I'll be back because I'm addicted. I don't use Facebook or Twitter - this is it for me.

Answer 4

There is a Stack Overflow technology that allows you to write good answers.

The Featured Tab

Go for the questions with an active bounty. A question only is eligible for bounty after 2 days, and that means the fastest gun is of no use here. Those are usually harder questions and sometimes they are rather interesting.

The situation is different for the featured questions:

• The questions usually involve a certain level of difficulty that cannot be solved in 5 minutes
• People usually take a longer time to read through both the questions and answers, since if they find the question interesting, they will take the time to look for the best answer
• The OP would definitely notice your answer if he/she is the one who place the bounty
• The improved traffic increases your answer's exposure
• The bounty gives you additional incentive to write an in-depth and canonical answer
• And so the answers tend to be longer

The home page is the playground for the not-so-knowledgeable-yet-eager-to-answer users, while the featured tab is for the experts. Pick the technology you like.

Answer 5

Currently, answering questions is incentivized over closing, as you can earn reputation for the former, but not for the latter. This lead to users answering, when they should close. If we incentivize closing over answering, this would lead to users closing when they should be answering. I think rewarding users for finding duplicates is a good idea, but it won't be easy to implement and safeguard against abuse. A first step would be to try a conservative rewarding scheme, e.g. only reward the first user to choose the exact duplicate, and take away the reward if the closure is reversed.

The more important point in my opinion is making it easier to find the ideal duplicate with the canonical and extensive answer. Some ideas I read about that essentially establish a second voting system, letting people vote for the canonical answer. Adding a second voting system is problematic and I'm not sure if the gains would justify it, but I could see it work if the parameters of the voting system would be sufficiently different. One could e.g. only allow users with at least a tag badge in the topic to vote for canonical answers, and one could severly limit the number of such votes a user has.

The second way to identify canonical answers is to look up how often they have been used as a duplicate target. I prefer this method, as it is far less complicated, but this doesn't ensure that the most useful answer is actually in the target question. To fix this we would need some large-scale tools to modify the duplicate network. Such tools would e.g. automatically collapse duplicate chains, allow mods to merge duplicate trees and similar actions.

Another idea would be to allow something like composite answers, consisting of a canonical and extensive post, with an introduction written by another user that explains how this solution works in the specific case of the asker. This would be more an idea for near duplicates and not exact duplicates, and I'm not sure if it would be all that useful.

I think this is a very hard problem to solve, but it is certainly worth trying.

Answer 6

I sometimes feel like a good answer is like "casting pearls before swine". I don't focus on my SO rep but it does sting a little when I spend an hour on an answer and it isn't even read (let alone upvoted).

Simple Questions

To generalize, SO's strong point is concise questions with the quickest answer of reasonable quality usually rising to the top.

But—even as a reasonably seasoned professional—that doesn't have to be a bad thing. I find great value in SO answers whenever I'm learning a new technology. And in our business, that's quite frequently.

Fastest Gun in the West

I'll be honest; many days I don't enjoy the FGITW scene and instead focus on cleanup tasks and constructive comments. But when I do decide to try to be that "quickest answer of reasonable quality" it helps me hone my efficiency and self-critique. It makes me read specs more than ever before because someone who knows the correct answer is usually reading, even if they aren't writing an answer.

In the couple of years I've spent on SO, my own base of resources has increased. Yes, SO doesn't have great tools for organizing reference material, but that doesn't stop me from building my own library.

Bottom line: learning to give a rapid, readable, accurate, sourced answer is not a bad thing.

"Bad" Questions

I don't need to enumerate all the types of poor questions which can be asked. I bristle whenever I see "Give me th codez" questions. But assuming the question is asked in sincerity and contains something technical (a link, some code, even a concept) I mentally take note. I may subsequently close the question if it doesn't fit SO, but I take note.

A week ago I was completely stuck on an issue pertaining to cross-device video support. I searched everywhere. In the end, I found the hint to my answer in the source code of a low-view question. The question had nothing to do with my problem. The tags probably weren't great. I don't remember if there was even a good answer. But seeing someone else's code led me directly to the answer I needed.

All said, I find sufficient value in the time I spend on SO.

Rewarding Appropriate, Quality Answers

I recently noticed a simple and mediocre old answer of mine was receiving new attention. This bugged me for a few days until I rewrote the whole thing (note the edit history).

It's still an appropriate answer to the question, i.e. I didn't try to turn it into a blog post or an editorial.

1. I personally received value by writing it. I learned a few things.
2. I can easily find it using the search tools. No complaints there.
3. No one has upvoted it since my edit; perhaps that's because people would rather have the one line of inferior code.
4. No one knowledgeable has commented affirming/disproving my response.

Point #3 bothers me a bit, but it's to be expected (and may just be happenstance).

Point #4 would be extremely useful to me. This is something we could improve; one line of encouragement (or criticism) goes a long way.

Answer 7

It seems to me that this jeremiad is a result of answering bad questions.

Good questions on Stack Overflow are concrete. They don't usually call for lengthy answers. They don't call for recommendations. So they don't offer surface area for answers in which the writer will "tell you to use [some technology] which they never have laid their hands upon".

It's important to be careful and thoughtful when writing an answer, but this is not the place for a detailed, Talmudic, commentary on anything.

Instead of taking on the role of the prophet crying in the wilderness, you would be happier if you took your writing skills elsewhere. If you really are an expert with lengthy analysis to offer on technology topics, people will read your blog. Or even your book. Leave these tar-baby questions to those who can't resist the urge to answer them, whilst some of the rest of us attempt to close them.

Answer 8

Perception

Today I had some time and felt like writing a relatively good answer on Stack Overflow; explaining both the underlying mechanisms and good practices, unlike usual declarative answers in a "do that" style.

That's awesome! I really enjoy it when I come across a question, new or old, where I have some particular expertise and I have an opportunity to share that expertise.

It'll bring me nothing but 3-5 occasional views, and - if I get lucky - some ignorant comments. ... It's a waste of time to spend so much effort to answer one question...

I'm sorry you feel that way. Looking at your account, I see that you've had several answers earn you the Necromancer - this is only given when you answered an old question and it gained at least five upvotes. Other badges such as Guru, Enlightened, Tumbleweed, Good Answer, Great Answer, Revival, and a multitude of Nice Answer show that your work is very much appreciated by the community.

Your most highly voted answer has 280 votes, and was posted a full three years after the question, fighting it's way near the top of the answers to be fourth out of 113 answers.

Efficiency

The current policy is driving knowledgeable and educational users away, yet it attracts inexperienced ones in numbers.

This is not a policy issue. This is an issue of too much popularity. There are millions of programmers who use this site, from beginner to expert, and it receives a lot of questions.

The problem isn't that answering with comprehensive answers isn't incentivized enough, the problem is that there's simply too much to answer, that answering one question comprehensively isn't as useful to either the questioners or the answerers as answering 20 questions quickly.

You speak of efficiency, but the reality is that if there's only a limited amount of experts to answer questions, and an unlimited amount of questions, the most efficient way to help the most people is to provide shorter, but still useful, answers than to provide comprehensive answers which limit the overall answer rate.

You further suggest that the answers are getting worse as a result - but if you peruse your own area of expertise you'll find that bad answers are penalized through normal voting.

Choose your own adventure

There are many ways to participate in Stack Overflow, and you don't have to choose one and stick to it, either.

Melee/warrior/healer

You can go for quick and dirty answers - this is fun and interesting, it's like a race and I've found that when I do this I end up learning more than I anticipated as I read other answers to the questions I've tried answering. Sometimes my answer is a hit, other times I'm schooled because I missed a better or more obvious solution.

I rarely come away from a session of quick and dirty empty handed. I almost always learn from it, and exposing myself to a variety of problems has helped me hone my own problem solving skills.

Get down in the trenches. Try to discern the real problem the OP is having from their terrible communication skills, and, like a field medic, try to provide the immediate, urgent care that is needed, or that will provoke a reaction that will tell you what the real problem is.

Note that you often aren't fighting the battle for them, or even with them - you're just trying to help them survive the encounter, and hopefully come out better for the experience.

Air support/strategist/educator/lawyer

You might want to provide a canonical, lengthy post to a particularly good question. Some lawyers looking to challenge a specific law, will search for a specific type of case that proves/disproves the law's validity. In this case you won't just answer any old PHP/MySQL injection attack, but you'll search for an example that is clear, has nothing extraneous or confusing, and will benefit not just itself when answered, but many, many other similar questions.

Once found, you would spend a good deal of time expounding on the answer that provides significant value not just to the immediate problem, but to those seeking to fully understand the problem and all its layers.

Rather than live in the trenches, you prefer to look over the battlefield, find a skirmish where you feel your effort will be best put to use, then you engage in it. You can see that many little battles have similarities, and that "winning" this one part will solve many other parts without expending significant resources fighting each one individually.

Big game hunter

You prefer the hard problems. You don't want to plink round with small weapons, you are battle hardened and know you can handle some of the most difficult problems programmers encounter.

You might hone your skills in the "unanswered" tab, but you really enjoy sinking your teeth into the "featured" tab. These are questions people are having such a hard time with that they are willing to part with a portion of the hard-earned reputation in order to receive an excellent answer. A 500 rep bounty might represent 10-40 hours of work for some people, and being the expert you are, you might be able to resolve it within an hour or so.

You'll note that many others are also vying for the big game, but you relish the competition, and know that you have the time and skills to provide the canonical answer.

Communications

You enjoy clearing the air, and making other's jobs easier. Performing triage on incoming questions, editing questions to add clarity, and asking simple follow up questions in the comments are your forte.

You take great pride in keeping things clean, neat, and orderly, and have an extensive high-level knowledge of the topic and terminology. This enables you not only to clarify and translate posts into topics that are actionable by the experts, but provides the skills needed to find duplicates. Duplicates not only answer the original question faster than waiting for an answer, but keep the experts busier on questions which either don't already have duplicates, or are different enough that they demand a different answer.

Conclusion

There is no need to lock yourself into one "play style" nor to tell others that if they choose a particular style, they must not be experts. You can take part in several aspects of the site, and you won't always find the rewards are reputation - many people prefer badges as a sign of their contribution, and some just participate to satisfy their own internal needs.

The problem you describe points to a few issues with the site - mainly too many bad questions - but doesn't mean the site itself is bad. There are many thousands of people who enjoy it, are experts, and give good advice and answers to the millions of people who ask questions.

If none of the ways to participate sound engaging or fun to you, it may be that you can think of a way you could participate that would fulfill the need you are seeking to satisfy. It may be that you can participate in that way without any changes to the site. The few "classes" I've suggested above are hardly comprehensive. There are many people who use the site in very different ways.

This is a choose your own adventure game, and it's complex enough that there are many opportunities for participation. Find one that appeals to you, then ignore the rest. There's no need to feel abused because you aren't receiving reputation and badges equal to your perceived contribution. Do what feels right, and ignore the race if that is distracting for the joy you get out of participating.

Answer 9

While there are no guarantees, writing a better answer is often worthwhile, even if you are not the first poster.

This morning I posted a fifth answer to a question that eventually ended up with 6 answers. By the time I posted the OP had already accepted a FGITW answer with 5 upvotes. That FGITW answer quickly ran to 11 upvotes while mine crawled to second place with 5. Two hours later my answer passed the FGITW answer and had been accepted by the OP on a revisit to the question.

In about a month on SO as an active answerer for C# and SQL questions this has happened twice. Those two answers will probably continue to receive occasional upvotes for some time. They also remain the standard I strive to achieve with my answers (though some questions will not be worth the effort).

Though I don't really believe it is relevant, here is the link for those who are interested (and to prevent comments simply requesting it): Question on Null handling in VB.NET and C#

Answer 10

I feel very sad that I can only agree with almost everything you have said.

I have not been using Stack Overflow for very long time - I signed up only three years ago. Since I am a PHP developer I started to contribute to PHP tagged questions first and later I started to contribute to questions tagged OpenCart (which is an open source PHP based e-commerce system).

I loved contributing to the questions - writing comments, editing and mainly answering the questions. From the beginning I felt that questions were better, that answers were more appreciated and not only by the asker but by other community members as well (I mean that the answers were up-voted). But these days are gone for a long time now. I stopped contributing to PHP tagged questions a long time ago and am contributing only to those tagged by OpenCart now.

For the last year I more and more often see only very low-quality questions from which 90% deserves closing for "beloved yet gone" Off Topic reasons (e.g. lack of research done before asking, question not concerning programming problem and even must contain source code to identify the problem was reworked and shall be used only when asking for debugging help...). For the last year there is only 10% of questions (when speaking of the mentioned OpenCart tag) that can be directly answered and from those maybe only half could be answered with a good answer. And those answers are up-voted only sporadically (You can check at Stack Overflow in my account - now almost 50% of my answers have 0 votes). Nowadays I feel more like editing and voting for close only.

I can see that with each day the number of closed questions and the number of questions that are not answered (and probably never will be) is only increasing. I can see that questions are being asked and even answered not by professionals nor by enthusiasts but by complete novices! You can see it in more than half of the questions: "I am new to PHP ...", "I am new to MySQL ...", "I am new to OpenCart ...", "My PHP skills are limited ..." and so on. I can see that almost quarter of questions are of type "How to do this and that in OpenCart?" while asking about the administration or settings only which do not concern any programming at all.

I remember that about two years ago we had around 4k close-voted questions waiting in the review queues - now it's more than 11k! This itself proves everything...

But though I am feeling more and more frustrated about the situation I do not stop. I try my best to edit all questions and/or answers to make sense or look better, I vote for closing the questions that do not belong here, I comment on low quality questions or answers asking for better description, and I answer those questions I can answer (sadly not because my knowledge is limited but because of the reasons described above).

So, please, if we want to make Stack Overflow a good place we should not turn our backs to the community, but help more extensively...

The increasing flow of inexperienced users is not the problem - but the leave of experienced users no longer taking care is.

Answer 11

I think it's a mixed bag of blessings.

Bad

One one hand, as an answerer, I completely agree with you. Somebody, who was first answering the question "what i++ does?" receives huge amount of rep/views and me spending an hour writing good/full explanation of some more complex idea or concept gets one upvote after a month. And this is really frustrating.

Personally, I decided to play FGITW game for most of questions and I have couple of subjects which interest me a lot and where I am willing to invest time, even if just several persons will read this answer.

Good

My memory isn't that good. I quite often forget stuff like how to set upstream on git or what is the proper way to initialize MediaRecorder on Android. And what I find fascinating, that I can find answers on SO litetally within a minute.

So, quite often, I don't want long and length description, I don't want to read through half of a manual, just to find correct command line parameter and so on. So, quite often huge amount of simple questions/answers works for me well.

Thoughts

I believe it's very important what do you want to get from SO. In the case, if you want to win in rep game... yeah... fastest gun (with big amount of time on hands) wins.

If you want to teach people how to be better software developers. SO is definitely wrong format. You should try a blog.

If you want to help people. You ARE helping. Even though you aren't receiving as much rep as fast guns.

If you want to raise your profile. I think writing good answer is WAY-WAY more powerful than writing hundred bad answers. I have 4K on SO (which isn't too much) and I got dozens of quite interesting connections through it (while my profile view shows 300). I know people having 20K with thousands of profile view, who literally didn't add a connection since they started answering questions (because nobody is interested to know guy answering "i++" question). So, as result, good answers can give you some real rep (through connection) vs virtual real (on SO).

Answer 12

I partially agree with what you say, but I believe the problem and thus the solution lies elsewhere. For me the problem is that for the community all questions are equal. An upvote on a question is an upvote on a question no matter what the question is. So my top answer is answering How to reverse a C++ vector?, took me two minutes, and has 33 upvotes so far. On the other hand I have some answers that solve or explain way more complex things like one where I show the difference between a trie and a radix tree. Took me much longer and has 0 upvotes.

So a problem I see is that repution mechanism does not take into account the level of expertise required to answer a given question. If there was a mechanism to give higher reputation bonuses for more complex questions, the problem you have with "sprinters" would also be solved - they may answer as many as they want of the simple questions really fast, while you can take your time and answer a single more complex question while still gaining the same reputation bonus.

Answer 13

Not even trying to guess whether anyone of SE team have read this, I'd say they act like they agree with concerns expressed here. Let's see what happened after March '13 when this has been posted.

In the list below, dates that aren't linkified refer records in Recent feature changes log. Text in small font is my comments:

• 2014-07-01: Draft design for more aggressive auto-deletion published.
  _{Attempt to stop incentives for lousy answers to bad questions. Upvotes and accept from help vampire to stop guaranteeing rep from bad answers, dupe closures to stop this either.}

• 2014-06-06: Explicit close reason for questions seeking debugging help.
  _{Attempt to make it harder to post bad answers that merely translate compiler errors for help vampire.}

• 2014-05-27: On Stack Overflow, asking questions is now rate-limited to 1 per 90 minutes for users having less than 125 rep.
  _{Attempt to limit amount of low quality questions that attract bad answers.}

• 2014-05-13: Users with gold tag badge have binding votes to dupe close.
  _{Attempt to make it harder to post bad answers into blatant dupes of already answered questions.}

• 2014-05-13: Anti-Recidivism System Added.
  _{Attempt to limit askers of questions known to attract bad answers.}

• 2014-05-06: Users with a track record of poor questions or answers will now receive a warning when posting a new question or answer.
  _{Attempt to limit questions known to attract bad answers and users having history of posting bad answers.}

• 2014-04-22: The 10k flag queue has been removed. Not an Answer flags now feed into the Low Quality review queue.
  _{Attempt to involve 30,000 2Kers into deletion of bad answers.}

• 2014-03-14: Crusade against link-only answers.
  _{Attempt to delete bad answers that are barely more than a link. Per my observations, it also looks like some time after that, SE team started running scripts that feed link-only answers (old and new ones) into Low Quality queue.}

• 2014-02-28: Crusade to burn down close queue.
  _{Attempt to limit questions known to attract bad answers.}

• 2013-07-24: Very Low Quality flags automatically add the flagged post to the Low Quality review queue.
  _{Attempt to involve 30,000 2Kers into deletion of bad answers.}

• 2013-07-02: Flags to close a question are no longer shown in the moderator only queue. Feed into common close queue instead.
  _{Attempt to get help from 20,000 3Kers in closing questions known to attract bad answers.}

Over 10 actions in about a year, all look like trying to address concerns raised here. Not bad for a discussion that was initially shut down with 18 downvotes and closure.

Answer 14

A lot of this is explained by the Pareto Principle, which boils down for Stack Overflow to mean that 20% of the answers are of higher quality than the other 80%. That's a rule you'll never escape, you can only marginally improve the situation by improving the process and training users. Note, that that improvement also raises the whole bar, so that even lower quality answers on Stack Overflow are better than most forums. The reason Stack Overflow is so good for Q&A is because it has such a high level of competition, just like how all the track records starting getting broken when more athletes from around the world started to compete.

Still, I often get help from the lower quality answers, because they will stimulate my thinking and help lead me to the solution I'm looking for.

While Tag Wikis aren't full children yet, this feature-request would help that a lot.

Answer 15

Part of your question addresses the DRY precept. The development of a good answer represents an investment of time, and it would be a good return on that investment from the community perspective if the answer could be shared among many questioners who have related questions, without much added investment on your part.

The DRY precept surfaces in different ways at different levels of abstraction. At the level of data definition and database building, DRY surfaces as data normalization. At the level of OOP and data sharing among objects, DRY surfaces as classes that extend other classes. And so on.

The problem that is vexing you is, in part, a problem with the SO community and the SO infrastructure as it relates to DRY at the level of sharing skills and knowledge. For every class of questions, there is a "canonical answer" that would serve to answer not only the question at hand, but also future questions in the same class. In the best of all possible worlds, some of those future questions never have to be asked, because the would be questioner is guided to an existing question, and is able to get the needed answer from one of the responses.

Some questioners have the insight to be able to read a specific answer to a specific question, and (sort of) infer the canonical answer from the specific answer, and then apply the canonical answer to their own question. Other users have to be led by the hand to specific answer to their specific question. Most of us have this ability to a greater degree in some areas of computing, and to a lesser degree in other areas.

"Please give me th codez" is perhaps the ultimate in asking to be led by the hand. The other extreme is a formal course of learning, where the concepts are featured prominently, and are of very wide applicability. The examples are meant to flesh out the concepts, and to give the learner some practice at applying these concepts to specific cases. Homework problems provide further exercise at applying the recently learned concepts.

The amount of investment in developing a good formal course of learning is immense, and dwarfs the investment in coming up with a good answer in SO. The amount of investment in taking a good formal course of learning is also very large, and dwarfs the investment in asking a good question in SO, and getting some good answers quickly.

A good formal course of learning also uses the DRY precept, but at a different level of abstraction altogether.

Vaguely connected to the DRY concept is the idea of creating a compendium of human knowledge that can be easily searched for the answer to any given question. People have been trying to do this since the dawn of civilization. Before I found SO, I had found Wikipedia and Google. Before that, the world wide web, and before that, the internet. (I don't know what order things actually got developed in). And if you go way, way back, there was the library at Alexandria, before it got burninated.

Sometimes, the answer I've given that has been most valuable to the OP is not the answer itself, but the right search terms to use with Google. In those cases, the right search terms gave the OP a wealth of material on the web from which that person could pick and choose.

The right answer to this question seems to be: pick a strategy that works for you. My strategy works for me, even though it only gets me less than 10K rep points in 5 years. I deserve some upvotes that I do not get, but I also get some upvotes that I do not deserve. It balances out.

Answer 16

I'll admit it - I'm guilty of gamification myself. I mean, it's fun - to get the first answer..

I think a lot of us might feel a little uneasy about what you're saying, because it's arguing for ideals we do aspire to - to make the internet better, to provide quality questions & answers. But perhaps we play the gamification thing anyway.

But I'm hopeful that as the technology gets better & faster, we'll be able to somewhat filter out questions based on quality, and maybe even provide user-tailored material.

For example, if a user Wiz737 qualified for and wanted to receive only high-level questions/answers, he/she could do so. Not sure how this works and it's really vague in my mind. But I like the idea of separating the wheat from the chaff - somehow :\

Answer 17

The more experienced Stack Overflow users can usually fix problems themselves by referencing code from here or somewhere else, whereas the less experienced users need a more specific answer that pertains to exactly what they are asking.

For instance:

A new user wants to figure out how to write out a program that does math with a list of numbers and makes his life easier. He only knows Java, so he doesn't really know how (or doesn't even realize how simple it is to) notice that one was already written in C++.

Older users can reference a C++ project and fix their problems, study different practices to improve efficiency, etc. Personally I think the Stack Overflow community is very well balanced out between new and old users.

I do see a good idea in your question though. A lot of questions are very specific and could be grouped together. I think that there could be lists of some sort (not necessarily tags, but like an archive of questions) that resolve one big project. Almost like calling functions, each question could be considered a function and people could learn the basics through studying each individual question, but a more experienced user might want to see how it was all put together.

Answer 18

Stack Exchange sites have mostly simple questions, in a way that they can be answered relatively fast (on seconds scale) if you know the answer or even if you only approximately know the answer. Your answer will be visible fast, and you'll get the points for it.

It's not only the Fastest Gun in the West problem, but also a problem of the questions being simple. A thorough, extensive answer might very often be overkill.

Nevertheless one could do something by tweaking the gaming system present here. For example by allowing to suspend voting on answers for a certain period of time. The rationale behind would be that if you want really interesting answers, you probably are willing to give answerers some time to compose thoughtful answers. So it should be in the questioners' judgment what kind of answers he/she wants and any support for this (regarding the working of the site) would be nice.

Answer 19

You invested much effort in an answer to a question, and now you want your investment to generate a be worthwhile. You need to amortise your effort. That is, you need your answer to be the answer to numerous questions, including questions that visitors ask themselves but do not result in them posting an answer. You need to drive traffic to your answer.

You should consider curating the question you answered, and all related questions.

Edit the question you answered, and especially its title, so it will be clear to visitors that have a similar problem that it is similar to theirs. Visitors are then more likely to see your answer.

Examine all the listed linked and related questions of the question you answered. Vote to close and down vote bad questions. Edit the titles of questions that have misleading or overly general titles. This will make the question you answered stand out as the superior question it is (you didn't put all that effort into writing a bad question, right?).

Search for duplicates of the question you answered. Vote to close the duplicates you find, citing the question you answered as the question they are duplicates of. You will probably come across questions that are similar but not actually duplicates. Add a comment to those questions, noting the relevance of the question you answered, with a link to the question you answered. This will create links to the question you answered, which will drive traffic there. Some of the questions you find will be crap. Give them the down votes they deserve, and consider voting to close for a reason other than duplicate, so the crap is buried or deleted, making the good question you answered stand out above the dross.

Be on the look out for new questions that are duplicates of, or similar to, the question you answered. Down vote and/or vote to close all such new questions that deserve down votes and/or closure. That is, treat these new questions just like the old questions you examined previously.