I often ask a question on Google, get taken to a Stack Exchange site, and find that yes, they have my question, but some moderator closed it before it got answered. SE got my traffic, but I didn't get my answer. Isn't that broken?
Roughly, there are two camps. Those who say that any well-formed, non-dup question of interest should be let in and those who think it's a stop of last-resort after checking all conceivable documentation and scouring the internet.
I believe the latter viewpoint is pretty run-of-the-mill developer machismo, and it makes me chuckle every time the first google hit directs me to a SO post with 1000 upvotes that has been closed as "off-topic".
It seems self-evident to me that in the year 2015, the internet has allowed me to make more valuable use of my time quickly searching for answers rather than scouring documentation. Docs have a time and a place, but are not needed for quick dips into the myriad techs you need to use to write a typical app these days.
What I would think would be better is for questions that receive a lot of upvotes after they are closed, that the people who flagged the question have reduced capability to flag down questions in the future.
I've some sympathy for that.
What do you think happens when the two are rewarded equally, but the latter can be done much more frequently per unit of time?
I don't think they're mutually exclusive - there are multiple reward structures for multiple types of players. Both should be encouraged.
Further, there's the question of how much effort the questioner should be expected to put in. In a great many technical question fora, the experience is that many noobs make no attempt to search out answers.
Heck, even if the site was only for "novel" questions, according to your logic, only the most important unanswered question should be visible, until it is answered. Otherwise people might be incentivized to answer the second most important question instead.
So, yes. Internet points.
I'm not suggesting we leave obviously terrible questions open, it's the borderline posts that get me. Interesting questions which may not have a black and white answer, but would still encourage discussion/opinions which would be useful to people down the road. I know "it's not a discussion site", but sometimes there are gray areas and it's useful to have the opinions of experts on the pros and cons of certain approaches/technologies/whatever.
I suspect that SOs answer to your need is "Go use a different site". At some point, that becomes a reasonable answer.
Just today I have probably worked with at least 20 code libraries with their own documentation, and most of those libraries will be replaced within 3 years from now.
I think reading the documentation for everything that a webdev works with now could be a full time job in itself.
This is wrong in multiple dimensions and [edit: the next sentence which i didn't quote is] flamebait.
One camp believes in following these guidelines to a fault. The other camp thinks some of the rules are counterproductive to SO's principles.
Source: http://stackoverflow.com/help, the guidelines are clearly outlined.
For example, where is the guideline that answers this without ambiguity?
http://meta.stackoverflow.com/questions/261592/how-much-rese...
I don't understand why I have to click another link to go to another site that doesn't have the same context and awareness of where I just came from.
In my experience, this is completely false, it should be reworded as "Questions that moderators like to remove on a whim".
The fact that a moderator (or group of moderators) don't find a question valuable doesn't mean that it is not valuable for the rest of the world, and this applies specifically with questions about recommendations.
I have gotten a _lot_ of value from closed SO questions that were high ranked in my Google search results.
I don't know what that says about SO's flagging system, but it is what it is, I guess.
Too bad for you, it just takes an overzealous moderator to remove a closed question that many people find valuable.
Your best bet is to bookmark the question and trigger archive.org's crawler via the "Save Page Now" form: https://archive.org/web/
Also in the rapidly changing JS world they don't reference the version consistently. Code that worked fine on version 1 doesn't work six months later on version 1.1, if I knew the version it would save a lot of time.
The problem is that you can't boil down development to very focussed questions with very focussed answers.
A lot of answers are experience and opinion - and if someone experienced is giving an opinion that's usually useful in itself, dammit, even if someone else with experience disagrees.
'Is React better than Angular?' may seem pointlessly open, but if you get answers from two camps who both know what they're talking about, it's possible to learn a lot from both.
It's unbelievably annoying when this happens. I have a number of bookmarks to incredibly insightful posts, whose usefulness has been corroborated by literally hundreds of people...all closed because they are deemed "off-topic" or some such by an admin.
My theory is that SO was essentially "done" years ago. But people are still on staff and needed something to do. Thus, unrelenting navel gazing ensued.
Imagine the question : "What is the best templating engine to use for Django?" This question doesn't have an answer, because it's so dependent on use case (even that statement is controversial). So if it's on SO and not closed, you'll get a bunch of answers that are contradictory, and people flaming in the comments.
I think the main objective for the SO team is to have people land on the site and _have their answer right there_. By removing possibilities of these off-topic questions, you're less likely to land on answers that help your specific thing.
End result is "when I land on SO I get an answer every time" instead of "when I land on SO sometimes I get a flame war". For branding, at least, I think it works in their favor to close those questions
e.g. http://stackoverflow.com/questions/257712/best-web-applicati...
or less seriously:
http://stackoverflow.com/questions/184618/what-is-the-best-c...
Also, smart, general questions are not the only questions to be closed: I got one closed on networking because the problem was between two company networks, not inside. There is an obvious reason for a rule about company networks and that is to keep hone networking out, not because we need another subsection for WANs, no?
how to push an item into an array in Java?
The obvious response is "lol don't you know how to read the doc"?
Yet the view count is over 100,000 and counting, with people commenting how much they appreciate finding the answer immediately.
So that's an example of a good question, actually. And an answer "read the docs" would be down-voted as unhelpful and quickly disappear.
The remaining comments are just fluff, such as "thanks", which can more appropriately (for the site) be expressed as an upvote or accept. Or they're asking a new question, in which case the parent's comment applies.
I listened to Joel and Jeff talk about it in their early podcasts when they were first building the site. To be the definitive answer for a question requires that the answer be able to change and evolve over time as new information becomes available.
Sometimes, I also see threads closed as homework questions, but those tend to have several duplicates which do have the answer in them, so it has so far not bothered me.
This is exactly what I don't understand. If a question is borderline, but obviously useful, why not err on the side of allowing it to stand?
>Sometimes, I also see threads closed as homework questions, but those tend to have several duplicates which do have the answer in them, so it has so far not bothered me.
This doesn't happen anymore. The "Homework" tag was removed and is no longer a valid reason to close a question. That doesn't mean these questions don't get closed using other rationals of course.
1: There is no definitive answer to it. At best there is best practice, and at worst there is flame wars. If Moderators tried to distinguish what topic could become flame wars, you tend to end up with very biased moderation.
2: Recommendations are much more time-sensitive than purely technical questions. The best framework in 2005 might be abandon in 2015, which can mislead readers. As a reader, I know this and always check the post date, but from a UI perspective, this key component is not obvious in SO.
3: Flame wars are destructive, and SE has historically tried to avoid them. Questions like "should I pick GPL or BSD?", "Emacs or VI?", and "Python or ruby?" tend to be cut down before they cause problem in the SE community. I wish sometimes that HN would follow suit and do the same.
This is why you can edit a question/answer, to update the information.
But now, I feel like my question is probably going to be flagged, so why bother.
I've said it before: SO is useful for the first 5 years of a software engineer's life. After that, it's occasionally useful when you need to pick up a new technology. The questions that are hard, for real senior software engineers, are rarely found on SE sites. Programmers.SE was supposed to fix it, but it hasn't worked, because of the same moderation.
This is a moderation problem, but the solution chosen was the easy solution: just don't allow hard questions because hard questions beget arguments. Instead, requiring people to have civil discourse, perhaps even requiring that people show they can have civil discourse before allowing them to participate, is much, much more difficult.
I'm not sure SE could survive the second type of moderation. But by not having it, it certainly makes it usefulness to somebody like me, with many years of experience learning and using various technologies, near zero.
1. http://programmers.stackexchange.com/questions/181342/r-vs-p... [closed as off topic]
That's what I find in my use and I don't think that's a problem. I think the harder to answer questions should be on their own Stack Exchange system. Moderation on that system would have different rules. I think if you keep it separate the second type of moderation you describe could work, however mixing the two types of moderation with Stack Overflow would lead to more problems imo.
They did open Software Recommendations Stack Exchange.
Closed answers get deleted in some regular interval. If they for some weird reason aren't, you could consider flagging them, there is always a response to that.
Although the question was framed badly it could have simply been edited to be more clear about what it was asking and what the answers are answering. It was helpful to me and I'm sure it's been helpful to others as well.
It's easy to assume that every reader is able to look past a poorly-written, half-implicit question and suss out the underlying need - but that's not the case. If you're able to understand a question and it's clear that others are not, don't hesitate to share your insights - they can make the difference between a question being downvoted and deleted and one that goes on to help many others.
I don't have an SO example on hand, but consider this Sci-Fi StackExchange question:
http://scifi.stackexchange.com/questions/79363/why-was-the-r...
The question is "Why does Sauron need the Ring to control Middle-Earth?" The correct answer is that he doesn't; he can control Middle-Earth without it, his victory is inevitable, but needs it to remove the only possible method his enemies have of overcoming him. After being correctly answered, the question was flagged as a duplicate of "What was the point of Sauron making the One Ring?" which is a very different question and comes with different answers.
This is, in my experience, extremely common across the StackExchange network. In that example, the "duplicate" question was answered before being closed, but that's not always the case.
Some examples:
http://stackoverflow.com/questions/21044302/make-ajax-loaded...
http://stackoverflow.com/questions/26298780/how-to-make-ajax...
http://stackoverflow.com/questions/14887710/ajax-deep-linkin...
http://stackoverflow.com/questions/21785074/seo-with-single-...
The comment to justify why the question is closed is always the same: "This question appears to be off-topic because it is about SEO".
FWIW, legit SEO questions do have a home on SE: http://webmasters.stackexchange.com/questions/tagged/seo
I do not follow you.
> FWIW, legit SEO questions do have a home on SE: http://webmasters.stackexchange.com/questions/tagged/seo
I know that Webmasters SE is supposedly the right place to ask questions about crawling issues and Ajax, but in fact you will find more questions on SO (see links below) because this is mainly related to technical issues with JavaScript.
234 questions tagged http://stackoverflow.com/questions/tagged/seo+ajax?sort=newe...
74 questions tagged http://webmasters.stackexchange.com/questions/tagged/seo+aja...
"Is there any way to dynamic generate content without affecting search engine indexing and ranking"
"Is there anything I can do to make google crawl my website"
"Is it safe to pre-load 4 website into one page without search engines interpreting it as misleading"
If you are a programmer who do not work at google, it is impossible to answer. If a co-worker gave me any of those questions, I would answer it by giving them the link to google's recommendation page about SEO, while warning that google is a private company that do not disclose their methods in indexing website. I can guess and speculate how they think or what their intention are, but any advice would only reflect how I think it works today, and google may at any time change how they index websites.
[1] https://developers.google.com/webmasters/ajax-crawling/docs/...
And I consider Google+SO the best programming tool ever created.
wrt Closed Questions, simple heuristic - if the page has lots of visits, lots of upvotes and is high in google for common searches do not close the topic. Clearly there exists a market for this question for many people, even if it boggles super-users' sensibilities.
The knowledge would certainly be useful to many people, but it is outside of SO's scope as it is defined.
[1] https://developers.google.com/webmasters/ajax-crawling/docs/...
(edit: typo)
http://stackoverflow.com/questions/19124445/how-to-download-...
http://stackoverflow.com/questions/19082860/where-can-i-down...
The question is off-topic, but as you can see it was still answered. Still, to solve that problem, a new site was opened per request of users: Software Recommendations Stack Exchange
How often is often in practice, though? This is a common objection, and I see it mentioned a lot by disgruntled commenters here. This is just another form of observational bias, and in reality the situaion described is pretty rare...
There are many times that SO and SE sites have been useful to me, both is finding answers to existing questions, asking questions that are then answered, and also providing answers myself. I agree that it can be infuriating to see a question that you want answered on the site, but it has been closed as 'off topic' or 'irrelevant' or the classic 'not suitable for a question-and-answer format site.' However, this has not happened to me much in the several years I have used the site(s) so I have to assume one of the following is true:
a. I only ask/search for the 'right' kinds of questions?
b. Closed questions are not that common?
Due to the annoyance factor, I think that b. is what is happening, and people just weight those occurrences much higher in their subconcious. That means they will recall them more readily, thus giving the false impression that it happens a lot.
This is actually a very common observational bias, which people should be more aware of.
- http://en.wikipedia.org/wiki/Availability_heuristic
The availability heuristic is the idea that if something can be readily recalled, it must be important, or at least more important than alternatives which are not as readily recalled. The easier it is to recall the consequences of something the greater those consequences are often perceived to be, and similarly the greater the consequences the easier somehing is to recall. This means strong negative associations with closed answers make those easy to remember, and people incorrectly estimate the frequency of this situation to be much greater than it actually is.
The overall quality of stack overflow has decreased dramatically over the past few years. These days, when I go there mostly I have the chance to help someone from India do his or her job and tell people to stop using obsolete database interfaces and leaving their code wide-open to SQL injection. It's not very rewarding, educational, useful or entertaining.
I sometimes feel as if there could be a thread titled 'SO pivots into airplanes' and the top comment would be about whether flight attendants would call some questions off topic.
Interestingly i have noticed that Quora has some nice thoughtful answers to tech questions, especially questions that have no shot in hell of being answered on SE.
Just to clarify, SE doesn't always have the question, but it always has the answer.
It must be related to the specific field you are searching for; older ones are usually complete. I wouldn't be surprised if a more casual programming field has a larger ratio of closed answers.
While I understand the intentions of the moderators, they seem to assess the value of a question for people at the moderator's own skill level than at the people who might be asking the question or seeking an answer to a particular question.
The silver lining for me is that at least they don't delete the questions and answers they mod down. I've found a lot of useful stuff in questions that the mods didn't like.
SE has good rules which keep questions relevant. I would say that if the question is closed and there are no existing duplicates, then the question itself doesn't have much value. ( There are of course outliers. )
There is a misunderstanding here, moderators on SE don't actually close most of the questions, normal users with privileges to do that, do so by voting. So such a question doesn't have value because users decided so, not moderators.
The last paragraph shares the misconception with the previous one. Users vote on and close question. Inf fact moderators are so rare on SE, that they only do critical things that should not be trusted to users, which do the rest.
Not fitting the guidelines of SO and not having much value are two very different things. Are you sure you meant the latter?
[1] http://stackoverflow.com/questions/4674609/looking-for-a-ric... (not my find, posted in another comment on this thread)
Having strict moderation standards is SO's business. I may or may not agree, but it seems to work for them.
Leaving closed questions, particularly those without answers, on their site sucking up google juice and deceptively appearing in my search results is scummy. If they don't want it on their site, they should ask google not to crawl it. And that goes squared when the question doesn't have an answer.
To a person _arriving at SO from Google_, "moderator" and "normal users with privileges" are basically the same thing, even though they may be technically different. And, those "normal users with privileges" tend to be overly aggressive when flagging questions as "off topic" or "opinion-based".
Googling 100 things (which is not a lot) is not likely to return more than 1 or 2 of these at most. It's rare that I hit this but I've seen it a dozen or so times at least. Most of the time it's a question that has been heavily upvoted with a heavily upvoted (and accepted) answer.