These replies infuriate me. The OP knows what his problem is and he's asking for an answer. He doesn't need a new problem to solve. He doesn't need to know what your opinion is about solving the problem. He needs a solution.
One example: I was searching for how to solve a RVM and rails configuration problem on OS X. I typed the error message from the console into google, and found a SO question that was exactly the problem I was having. But instead of providing an answer, the most upvoted comment was to uninstall everything and use homebrew because "you're doing it wrong" without homebrew.
That's nice and all, but I've had a rails and rvm setup on my laptop for 2 years without homebrew, and, like the OP, I don't have time to set that up now. I just want to fix this one configuration error and get back to work. I quickly went back to Google and found the solution on another rails forum.
Granted, this is a problem with a lot of internet forums, not just SO. But it drives me nuts.
After a long hiatus, I made a new account a couple weeks ago because I didn't remember my old account. Minutes after I posted it, I receive 2 "doing it wrong" answers, someone edits my title and tags, and then another person closed in minutes for being a duplicate when it wasn't.
As one of the first people to reach 10k points on SO
(I wrote the first SO point guide discussed on Meta) no case could be made that I didn't ask a proper question. I think there are too many people with 10k points now.
Stack Overflow has become a very toxic environment.
In your example, perhaps you didn't know about homebrew and this was going to save you a lot of time, well that's great. There can still be other answers that help your specific issue, but someone has thought outside the box and offered a completely different way of doing it.
For me the value of SO would be greatly reduced if people didn't do that.
However, it is very annoying when I ask a question and the only answers are some variant of, "you are doing it wrong." Multiple times I have had to write an explanation that is longer than my actual question to convince people that I actually do want an answer to the question I asked.
If you really think you understand my problem better than I do, based on a simplified explanation, awesome. I would love to hear a better way to go about what I am doing. But please, try to answer the actual question first (or at least say you don't know the answer). There is a slim chance the "better" way is not actually workable in the real world scenario.
When people ask questions online, far too often the person answering assumes that whoever asked the question must be a complete idiot. There are plenty of topics in which I would consider myself an expert, but I still get treated as if I know nothing about the subject. (I've noticed the problem also manifests itself in academia -- the more prestigious the academic institution, the more everyone assumes everyone else is "incompetent by default").
What irks me is when a google search lands me at SO only to have it marked as a duplicate with no links to the actual duplicate, or when the linked dup is dissimilar or unanswered.
The biggest problem is that some people treat it like a game of Jeopardy, except with no penalty for incorrect answers. This is especially true when the questions are dealing with weird corner cases in languages or libraries - it often looks like there is an obvious answer, but it isn't always so simple.
I used to take the time to get the code running on my machine (if it was simple enough), figure out what was wrong, fix it, and then take the time to try to educate the person asking, rather than simply telling them what code to paste in to fix it.
This takes time, and in while I was working on it, you'd have someone come in, and blurt out some guess as to why it wasn't working, then edit their answer a bunch of times and maybe eventually get it right. I don't know why, but it always bothered me to see someone who guessed around (and maybe looked at better answers in the meantime) getting credit for really sloppy work.
I guess I don't know how you'd fix that, and it doesn't seem to be a big enough of a problem to keep the site from being useful, but it is still kind of annoying.
I agree there's a problem, and used to participate anyway because more fun or interesting questions were getting asked early on; I never seem to see new ones when I peek in nowadays.
Added (heh): people also vote for sloppy answers just because they sound good. I don't know what to do about this -- perhaps you could try to estimate a voter's average vote quality, and weight them.
I take issue with his issue number one because I think the blame is placed on new people FAR too often online. The way new people are treated on pretty much every online community I've been a part of has been horrible, often downright pointlessly mean.
If we want SO askers to have good manners its our duty to explain our customs. When I answer a question from a new person I try to welcome them to SO or applaud their question (if novel), and if they worded it poorly I might answer what I think they meant and suggest better wordings to get help faster in the future.
I symapthize with the autor, but at the end of the day I choose to stay on SO in spite of all its woes because the gratitude of the answerers alone is more than enough for me.
I tend to go for the computer vision and finance questions since thats what interests me... but eh I can't participate due to those reasons anymore. I just don't care, and its not worth my time. I will however still use it to ask questions.
My other problem at the moment is that I am working on some objective-c stuff and SO is mostly useless for finding answers. Being new to the language and the APIs, I am sometimes lost and trying to get my bearings. A few questions I've submitted have either been closed or ignored. Only one really got a good answer, and when I tracked down the author I found he is very skilled and helps out lots of people. But there only seem to be a few people like him.
And by the way, if you're in SF and know objective-c, I'll buy you a couple beers in exchange for walking me through a couple things.
I'm finding I'm starting to go back to mailing list for niche programming topics (e.g., a specific framework or library) because nothing beats getting an answer from a top contributor or owner. Most people don't sit on StackOverflow all day waiting for questions, but they will respond to e-mails that pop up on their mailing list.
I was a frequent contributor to StackOverflow  but have largely stopped for a number of reasons, the most important of which is I got a new job that took up much more of my time.
But another reason is that for me, as a (then) frequent answerer, it got a whole lot less interesting. This was due to two factors:
1. A lot of the low-hanging fruit had been answered so the questions became increasingly esoteric such that you were less likely to simply know an answer and had to spend more time researching. That extra time meant you were also less rewarded for the answer because less people were in a position to state that it was correct or not (if you consider karma a "reward"); and
2. The ceaseless campaign against "interesting" questions due to increasing closure due to "subjective and argumentative" and the fragmentation of SO into the many StackExchange sites (causing a lot of questions to be migrated).
There are three basic errors that Joel has (and Jeff had) made (IMHO):
1. Over-emphasis on editing.
In my mind there are three groups: askers, answerers and editors. Jeff & Joel made statements about editors are important and how editing is super-important, basically trying o elevate it to the same level as answering questions.
This is a problem.
Editors are the bureaucrats of the StackOverflow ecosystem. The bureaucracy is expanding to meet the needs of the expanding bureaucracy . The more editors you have the less each has to do. Rather than doing less, the kinds of people attracted to this kind of function prefer to simply create work for themselves.
What's more, from Meta StackOverflow, from interacting from the people who edit a lot (and answer very little if anything), this simply reinforced my view: these are the kinds of people who destroy communities.
Those who can, answer. Those who can't, edit.
I've seen many spurious edits to many of my higher voted answers. Some capitalize something. Others come along and uncapitalize it. I've seen people come and add lines to my answers saying I stole it from someone else (seriously).
The net result is virtually any highly voted answer I have has been edited into community wiki oblivion. That creates a strong disincentive for me to spend time coming up with a good answer: I'll basically get limited credit from it and then have to watch as wanna-be editors essentially vandalize it.
The other thing these people do is close questions on the drop of a hat. They, as a group, tend to have an incredibly festidious nature when it comes to the enforcement of rules. They constantly seek some purer, higher standard and don't quite get that those rules are guidelines that are a means to an end and not an end in themselves.
2. "Subjective and argumentative".
Jeff and Joel from the outset wanted to prevent flame war type questions, the kind of questions that have no definite answer. Questions like "Is Ruby better than Python?" That's fine but it's been taken too far, in part by the very editors I previously mentioned.
Questions like "Should I use Angular.js or Ember.js for developing a CRUD-type Web application?" would be shut down in a minute. But the extra context matters. You can answer that question by giving a list of comparative advantages without being necessarily biased or inflammatory. That's actually interesting content, particularly if you're trying to decide between a number of new technologies, languages, etc. But alas SO seems keen on shutting that down; and
There is another thread today about the difficulty of music classification . The same goes for Q&A. Hierarchical classification schemes are too limited for Q&A. There are SO questions that have bounced around between a number of SE sites for this reason.
Just look at the Stack Exchange sites . If I want to ask a question about being a programmer do I ask on OnStartups, on the Programmer SE or elsewhere?
The pragmatic answer is that there are some question that naturally fit on several sites. Yet the hierarchical pigeon-holing with esoteric and often subjective rules means it's harder to find content, it's hard to find where to place content and newbies inevitably get chastised for posting on the wrong site.
There is a reason tagging exists and is successful. Describe the traits of the question or, better yet, figure it out from the content, and show it when it's relevant to someone. Don't make me hunt across sites for it. That's a ridiculous solution.
I predict you'll see the rise of more tagging-oriented (at a higher level than say Java or C++) and automatic classification Q&A sites in the future.
With that out of my system, let me address som eof the points the OP specifically raised:
> The Eternal September Issue.
This one annoys me. And I don't mean new people. The negative reaction people have to them (eg ). When a community starts chastising newbies, that reflects badly on the community, not the newbies.
> Down voting as a means of closing a question.
The one part of downvoting I don't like is people use it as a means of saying "I disagree with this" (on purely subjective grounds), which is not the intended purpose. That problem seems to be nearly universal with voting systems (even here). The one good thing SO does is "charge" you for downvotes. That alone stops it being a huge problem (IMHO).
> This is another one of the odd cases on StackOverflow. A few of the “Exact Duplicate” questions are not duplicates due to minor, but important, differences.
True. The problem here again is that you have editors deciding to close things that they don't necessarily know anything about. The same problem infests Wikipedia (deleting articles on "notability" grounds).
> The value of reputation: After the global recalculation, the site’s creators made a bold statement that participation is not valued on the site.
I actually don't know what this is referring to but then again I've been largely inactive. There have been several recalculations though (eg question upvotes from 10 to 5 karma). I don't really have a problem with this. If you're too obsessed with your karma, you're focusing on the wrong things. And I like that the same rules apply to everyone (eg it's better than question upvotes before X are 10, after X are 5).
I worried when SO took VC money that they were going to turn something that is very successful in one segment and ruin it with attempts at making a general Q&A platform. 2+ years on I'm still failing to see traction in the SE sites beyond SO. This may yet still become a problem.
Perhaps the simplest answer here is that SO isn't a "community" as such. It went with the Q&A format (over, say, forums) to discourage discussion. People often criticized it for making discussion hard when that was kinda the point.
I certainly never went to SO to hang out. Some do I'm sure. The focus is (and should be IMHO) on the content not the community.
The other day I was doing some searching on the use of express in node.js. I think I found about 10 relevant and useful threads of which I think only 1 was still open. All the other threads I actually found very useful had been closed by some moderator. Subjectively, I think this is a very typical ratio.
I just don't get it quite frankly.
It didn't leave me with a great impression of the community TBH, and felt overly draconian. Of course I didn't make a big scene about it, and will still contribute now and then after having adapted, but I think that's a shame in a sense, because often broad and open questions generate a lot of useful discussion.
Almost any technology answer usually has an "it depends" in it somewhere and that of course needs a discussion in order to fully explore the answer.
I've said it before, I'll say it again -- the creators of StackOverflow don't know what made the site successful in the first place.
SO is successful because it's an experts-exchange.com clone without the pushy subscription model plus a few social voting features added in.
The problem also is that the site was essentially complete within the first year. But there are still people on staff and they took a bunch of VC money. The people on staff need to do something and the VC money needs to be spent.
So of course they start meddling with their already completed product. Editing is now seen as an Important Thing and many red herrings like "keeping the site from becoming Reddit" are trumpeted in order to justify this meddling.
Nonsense like "Gorilla's vs Sharks" are trumpeted in order to justify this meddling.
And even worse than the meddling on the new stuff, they go back and close old questions, even ones that have been quantifiably validated by the votes of hundreds of people. I'd actually bookmarked many of them, only to find them mysteriously closed months later.
Holy shit, are you fucking kidding me, guys?
There have been a couple of suggestions in science fiction that would be interesting to see someone try. Heinlein suggested a special branch of government whose purpose was to repeal laws and regulations promulgated by the other branches. Frank Herbert went to an even greater extreme, by having an quasi-official "Bureau of Sabotage".
Perhaps online communities should have a separate group charged with reversing the decisions of editors/moderators/whatever they're called, with no discussion or appeal permitted.
So, yes, a system of meta-moderation, and preferably one that is mechanical and resistant to being gamed, is needed. As fascinating as it is to see emergent cultures in Wikipedia editing and reddit moderation, it ends up being dysfunctional, and that doesn't serve the goals of the communities from which those sites derive their value.
I like basketball. But I wouldn't expect stackoverflow to allow basketball questions.
This isn't a judgment on the value of discussions... but there are other sites for that.
Communities of Practice (Lave, Wenger) are held together by sharing a common repertoire of practices, and opinions. Newcomers may want to explore these shared opinions. SE is obviously not a place where they can do that, so I end up agreeing with your final sentence.
What are the drawbacks of Python?
Just because you like something doesn't mean that it has to featured on Stack Overflow. Same as politics discussion on HN, many might want to debate Romney vs. Obama with the HN folks here, that doesn't mean HN has to tolerate that and turn into another Reddit.
I don't understand all the complaints about having discussions that can easily degenerate into flamewars with thousands of "answers" on Stackoverflow. If it's moved to a Stackexchange with less traffic, go hang out there and make it a better place.
>They aren't the ones creating a great site. Those that ask and answer questions are.
Sure. Look at a site where there is no moderation. http://answers.yahoo.com/
My experience with online communities tells me that the slippery slope in this case is very real. When new and immature folks generate content with back and forth among themselves(voting each others' comments up), the old timers simply pack up and leave since the site doesn't interest them anymore. Happened to Reddit right before my eyes.
So why not use programmers.stackexchange.com for such discussions? The more people use it, the more active it gets. Stackoverflow has already made their opinion clear on this.
The global recalc happened arround Feb10. They applied the new policy retroactively. My reaction was ffs.. how can I actually get credit for the participation. Theres no value in supporting the site other than the reputation you have. The reputation should reflect on how helpful you are. The karma you get there is helpful to demonstrate to employers how well you can help others, thats a little difficult to explain how your score just got deflated.
There were users who were earning huge swatches of rep by asking hundreds of questions and giving little back in return.
I lost ~1k in rep back when I was >20k, and 1k back then (relatively speaking) was a fair old chunk. It was disappointing at the time but it was the right thing to do.
That decision told me... hey tomorrow they can just decide that all users with name x, y, or has numbers in the name can go bankrupt on reputation.
If anything many of the recent changes have been to make the site more inclusive such as suggested edits, better review processes and so on, so that users who are still on their training wheels on the way to the the 10k and 20k tools can practice "community moderation" under the guidance of more experienced users.
That's the impression I got listening to all the SO podcasts. A quick search found this article, "Stack Overflow Careers: Amplifying Your Awesome" http://www.codinghorror.com/blog/2009/11/stack-overflow-care... and this quote:
'If you're actively participating on Stack
Overflow, we now have another way to convert
those slices of effort into something that
actively furthers your professional goals
Stack Overflow Careers.' Coding Horror.
may/maynot & explainable. Would and employer look at a third party like that without looking closer at the candidate? Or are they using the suspect rep numbers as a basis of exclusion?
So, SO is turning into Wikipedia…?
Literally over 50% of the SO pages I end up on are closed questions. If I'm lucky, some kind soul got off a good buzzer-beater answer before the Content Police showed up.
I do the only thing I can. Upvote the question (which can still be done on closed questions), and tell the questioner it's good, and the moderator is being an ass. Which they usually are.
The kinds of questions that annoy me aren't "do my homework"; they're real questions that focus on specific issues that I'm encountering and wasting time on. Some brownie-point seeking moderator is engaged in exactly the type of behavior that other comments note.
I have not yet contributed, in no small part due to the above. I don't need SO karma to prove my knowledge in a given domain, and therefore my only reason to actively participate would be if I find it intrinsically rewarding - which does not currently appear to be the case.
"Standards are enforced extremely rigorously"
Do you see the link?
Many people claim that down-voting should only be used in very specific cases, but The Stack Overflow guideline seen while mousing over the down arrow is "This question does not show any research effort; it is unclear or not useful." I think that (especially the insufficient research part) is pretty broad.
I called him out on that insanity, he was a big member on CodeProject, as was I. His excuse for that "codeproject turned in a shithole, lets not make this place like that" [not exactly what he said but it sums up many of his responses].
I think the "we will teach you a lesson when you disagree" attitude is pretty juvenile. It also seems like it's bad for business. People won't ask questions so there will be fewer search results directing people to the site.
"@event_jr. Screw you. I'm looking for a quick elisp solution not your opinion. No one is doing my homework. I'm just getting a answer to something that I thought would be useful. Thanks correct. I don't want to put a lot of work into it. I have a Perl solution that'll do."
"I did zero, zippo. No work before asking."
So you know, you only had yourself to blame there.
At the moment, I'm pointing out the the juvenile "beat downs" that happen certainly don't make the site appealing to a certain group of people. I'm sure that I'm not the only person who has quit in frustration.
Btw, it's great that you found my deleted comments. Can you also include his snarky comment to which I'm responding?
That said, nowadays I mostly browse ux.stackexchange which isn't as nitpicky yet.
I can see why though, there are so many damn dumb questions, the moderators have to be sick of them by now.
It's a lot easier in the newer sites like DBA.
The answer they provided? "Try to send an email there and if it works, it is valid."
How is this supposed to help extract emails from text? Stack Overflow has become a hostile and unhelpful place.
I'm sure there are plenty of other cases of good moderation, but to me it highlights a serious problem with the site. One loose cannon (who I can only assume either was a badly written bot, didn't speak English, or was just trolling) and one overactive mod managed to ruin that question for everyone, and the rest of us had little power to really help.
I don't know what the answer to this is; it bears a lot of resemblance to Wikipedia, and they still struggle with it.
You know, there's a flag link under every post. If you thought a question was wrongly closed by the community or a mod then flag it and explain why. Nothing is irreversible.
I even have moderator privileges, but I don't have the time or desire to bother anymore...
I also started a discussion on meta-SO about "fix my code" questions with a concrete proposal , but didn't get anywhere. The most operational (and upvoted) answer/comment there was that the "close question" system was intended for this and that no other mechanism was needed. Well and nice, but maybe we'd then also need a filter that would filter-out questions with 2 or more votes to close. (You could still opt-in to see these questions, by they shouldn't be on the frontpage by default.)
Also: if I did mention support for each of the points whats stopping those examples from being edited, deleted, or someone from aruging "those are extreme cases"?
Surely there must be a way to capture the value-added for all those thousands of people in the reputation system? Many people get to these answers from google, so it sounds like a simple matter of pinging them to (sign-in) and upvote the answer that helped them out.
If, on the other hand, as the author says SO is no longer actually values reputation, and is screwing with their algorithms accordingly, that is a serious problem. Not sure what can be done about that.
This happens if you're already logged-in and got to the site through a Google search (although only under certain circumstances—I think you have to have not visited SO in the past few hours/days?)
> If, on the other hand, as the author says SO is no longer actually values reputation, and is screwing with their algorithms accordingly, that is a serious problem. Not sure what can be done about that.
The only changes to the reputation system that I can think of are:
- Making question upvotes worth +5 instead of +10 (March 2010: http://blog.stackoverflow.com/2011/06/optimizing-for-pearls-...)
- Making answers on downvotes free (June 2011: http://blog.stackoverflow.com/2011/06/optimizing-for-pearls-...)
Beyond that, you can now gain rep from suggesting edits (+2, max +1000 from suggested edits or when you gain editing privileges), but that's not a change so much as an addition to the system.
Can't think of another issue where the angst is so disproportionate to the underlying problem.
I know, duplicate posts take up system resources, and we all remember when Technet and Usenet were DoSed by one too many duplicate posts (or did that not ever happen?). But duplicate posts also make common issues /usability problems more visible, and illustrate how terrible every forum's search is.
And besides, we've been trying to stop duplicate questions in tech forums since what, the first BBSes in the 70s? It hasn't worked, the fight is not worth fighting.
I recently solved a tricky bug in some Java code that we eventually realized only happened after de/serialization upon server restart. Seeing no related questions on SO, I took the time to create a self-contained test case as part of a question and also to answer that question with a solution that showed how to use reflection to set the value of final fields.
For my trouble, several SO members trolled or ridiculed me. This doesn't exactly encourage future participation. My time is too important to waste it helping self-important, ungrateful jerks.
 nicholas.hauschild trolled me for answering my own question (deleted by Anna Lear), and you can see EJB's discouraging comments.
But don't be discouraged. Ignore the trolls, flag for mod attention and we'll sort out crap like that.
Personally, I think StackOverFlow isn't worth the effort. What people there don't get is that what's really important is simply getting great questions and great answers. That's it. Period!
As far as using the score to get a job, I don't buy it. If you're actually good at some topic you will be able to get a job regardless. If I was hiring someone who was pushing their super-high Stackoverflow score I would be suspicious why they spent so much time gaming a dumb score on a website instead of building cool stuff.
SO rep is just another small nugget of information that may be useful to potential employers, and if said employer is using the SO careers site, your Careers profile is linked to your SO profile page.
I mention in my "other activities" section of my CV that I participate in the site, but I certainly don't make a big deal about it, and especially the rep part.
I have to contradict the article's assertion that karma's were dropped by the rebalancing ... I've been amazed at how my better answers have continued to receive up-votes and I think my reputation is about 1000 points higher than it was when I last participated.