4 ms·

The article casts users as the problem: help vampires, noobs and reputation collectors. My strong opinion is that there are other serious issues that the artic

The article casts users as the problem: help vampires, noobs and reputation collectors. My strong opinion is that there are other serious issues that the article doesn't even touch on: 1) the silo model of creating dedicated SE sites. There should have been one SO for everything. Collaborative filtering etc would have been an infinitely better direction to go in. 2) the over-moderation by misguided 'defenders' who are too eager to mark something duplicate or move it to some dead SE site. Here's a group of users the article doesn't point fingers at, but which I believe is more of a problem than the noobs! 3) the shelf-life of answers. Almost all the accepted answers are now wrong. Frameworks change, features are added to languages, bugs get fixed. And woe betide anyone who asks a question afresh hoping to get a new and actually working solution... 4) that I'm ok with subjective questions. These ought be handled as such, rather than closed. There are a gazillion little ways to fix this and make SO relevant again, but I think the mgmt and 'community' are a big part of the problem, not solution :( (disclosure: I have well over 40K rep, most of it on SO)

endgame 10y ago
In addition, the reputation model is actually incentivising the wrong thing. Answering some simple question that many people see will give you an answer that gets upvoted to the double or even triple digits. Answering something hard and well might only get a handful of votes if it's a niche topic. (disclosure: 11K rep, mostly on tools like make, autofoo and cmake.)
- alixaxel 10y ago
  ^ This.
- alixaxel 10y ago
  I once thought of a potential solution where the number of reputation you would gain would be directly derived from the number of tag-experts who saw the question (and did not provide an answer). But maybe this would lead to it's own problems too.
  - Drdrdrq 10y ago
    Bounty somewhat alleviates this problem, but only partially... Maybe there should be an option to "tip" good answers, where (part of) the tip would be taken from your own score?
    - mafuy 10y ago
      I'm not sure what you mean, isn't this exactly what bounties on SE currently do?
willvarfar 10y ago
PS: anyone want to crack an interesting problem, and earn a bounty in the process? A few days left and not a single answer on my recent question here: http://crypto.stackexchange.com/questions/37102/culling-equivalent-enigma-keys-for-short-messages http://crypto.stackexchange.com/questions/37102/culling-equi...
Strilanc 10y ago
> There should have been one SO for everything. What could possibly go right about mixing Judaism questions and Christianity questions into the programming questions?
- willvarfar 10y ago
  Its something I've mulled for years ;) Here's about collaborative filtering: http://williamedwardscoder.tumblr.com/post/15581427232/self-organizing-reddit http://williamedwardscoder.tumblr.com/post/15581427232/self-... Here's just general SO discontent: http://williamedwardscoder.tumblr.com/post/25426541504/stackoverflow-unwinding http://williamedwardscoder.tumblr.com/post/25426541504/stack... and http://williamedwardscoder.tumblr.com/post/26580091073/stackoverflow-humour-underflow http://williamedwardscoder.tumblr.com/post/26580091073/stack... ;)
  - reitanqild 10y ago
    The collaborative filtering thing stands out as a really good idea to me. Anyone here who can say anything about the community size you would need for something like this to be viable?
- reitanqild 10y ago
  Ask reddit. On a more serious note: Stack Overflow started out as a technical site. Questions about programming often crosses the border to hardware or networking. They rarely however cross into theology. I'd be fine with opening a different site for theological questions as long as I could have all my technical questions in one place (and more importantly not having someone move them around.)
- borispavlovic 10y ago
  Is this a problem with wikipedia.org? Imagine both of these sets of questions tagged as `religion` and you have a common source for all people interested generally in any religion.
  - pwdisswordfish 10y ago
reitanqild 10y ago
My thoughts as well. Wish someone would create that site. Will be happy to try to contribute. There is so much SO got correct that I think just changing a few rules could work as long as we could manage to get the word out. As managed above silos probably makes more problems than they solve (partially because of the moderators but I guess the silos should go away anyway). Same goes for strict rules about what belongs (Case in point: question about inter-company WAN traffic being rejected on networking because site guidelines says something about "inside corporate networks"). Rules about what does not belong (acceptable content) should be OK as long as they don't prevent technical questions from being asked and answered. A way to deprecate answers (and a way for viewers to "flag" them as "still useful".) And yes, a way to handle subjective questions. (Quora already has a grip of this only their IMO insane policy of sharing my every click on the site made it unusable for me and a number of others.)
CannisterFlux 10y ago
On 2 and 3 - this is dead right. As an example of 3) Here's a question that has an accepted answers that is wrong http://stackoverflow.com/questions/7635013 http://stackoverflow.com/questions/7635013 - it's regarding x86, x86_64 and x32. Since x32 was new/unknown at the time, a lot of the answers say "x32 doesn't exist". And for 2) here's an answer http://stackoverflow.com/a/26522110 http://stackoverflow.com/a/26522110 that is/was true at the time of writing - a nasty little bug in the Android NDK causes math library functions to not get linked in on pre-lollipop devices if you state you are using android platform 21 and ndk 10c. Two rules lawyers are spouting off about how the poster should post a new question or leave a comment. They are in over their heads. The comments are out of context and come across as patronizing. It is also now an example of 3) since all the tech involved is pretty old in Android terms.
tlogan 10y ago
From the point of view of person who uses SO just to find solution, incorrect answers are the biggest problem. It seems that "you don't know what you don't know" does not compute with many people involved in answering SO. But that kinda make sense: at the beginning, SO was great because answers for noob questions where hard to find. At this point, there are so many other resources to get these noob questions answered or they are already answered. Now, people start asking hard stuff - but engineers (I intentionally do not say developers) who know answers on these questions are not interested in answering.
anexprogrammer 10y ago
I used to spend quite a bit of time on SO and OnStartups. I haven't contributed to any of their sites since the ridiculous deletion of OnStartups due to the opinion of one moderator (If I remember correctly, which is vague now). I don't care that I lost half my rep. Some people had put really lengthy, researched, answers and a valuable resource was just rm'd. There's far too many silos. Sure separate maths from games from Java. You have a tech question, is it webmaster, server fault, super user, or what? Seems like there's a very slim chance of getting it right first time. Sometimes it's the remit of several but none seem to want it. Or get marked as duplicate when it's fundamentally different to the marked duplicate. Some of the subjective questions have attracted the best, or most useful, answers. Even when they've survived years they still sometimes get deleted. Stack Exchange isn't worth the effort any more.
- DanielStraight 10y ago
  Wow... you're not kidding about the silos. I haven't used Stack Overflow much lately, so I thought it was still just Stack Overflow and Super User. Instead... Stack Overflow Server Fault Super User Web Applications Webmasters Game Development Tex - LaTeX Unix & Linux Stack Apps Ask Ubuntu Wordpress Development User Experience Theoretical Computer Science Programmers Information Security Database Administrators Code Review Programming Puzzles & Code Golf Drupal Answers SharePoint Software Quality Assurance & Testing Computational Science Mathematica Computer Science Raspberry Pi Network Engineering Open Data Software Recommendations Arduino Joomla Data Science Emacs Vi and Vim Open Source Computer Graphics Hardware Recommendations 3D Printing Retro Computing And I'm sure I missed some. Unbelievable amount of fragmentation. Anyone want to take a crack at explaining the difference between "Computer Science" and "Theoretical Computer Science"?
  - mafuy 10y ago
    I'm remotely involved there and my reply would be that TCS is research level while CS is not. There seems to be significant overlap in the questions, regardless.
danmaz74 10y ago
> make SO relevant again Curiosity: is there any more relevant alternative?
jlericson 10y ago
I sympathize with 2-4. Those things make it much harder to earn reputation by asking then it used to be. Fixing those problems requires fighting a bunch of institutional inertia, so it will require either time or some sort of judo move. But I don't understand how having separate sites relates. If you contribute to some of the smaller sites, I think you will find you'll get excellent answers from people more focused on the sort of problems the site covers. Frankly, I've seen DBA questions asked on SO that get short-sighted answers from programmers. Those questions should be asked of people who really know how a database works under the hood. I can see how collaborative filtering might help. But a big reason people have moved to dedicated sites as opposed to SO is that they want to have separate moderation policies for the questions they enjoy answering. (I've put a little bit of thought into the criteria for spinning off sites on meta.SO: http://meta.stackoverflow.com/questions/271989/does-it-pay-to-spin-off-sites http://meta.stackoverflow.com/questions/271989/does-it-pay-t...) (Disclosure: I'm a Community Manager hired by Stack Overflow.)

endgame 10y ago

In addition, the reputation model is actually incentivising the wrong thing. Answering some simple question that many people see will give you an answer that gets upvoted to the double or even triple digits. Answering something hard and well might only get a handful of votes if it's a niche topic. (disclosure: 11K rep, mostly on tools like make, autofoo and cmake.)

alixaxel 10y ago
^ This.
alixaxel 10y ago
I once thought of a potential solution where the number of reputation you would gain would be directly derived from the number of tag-experts who saw the question (and did not provide an answer). But maybe this would lead to it's own problems too.
- Drdrdrq 10y ago
  Bounty somewhat alleviates this problem, but only partially... Maybe there should be an option to "tip" good answers, where (part of) the tip would be taken from your own score?
  - mafuy 10y ago
    I'm not sure what you mean, isn't this exactly what bounties on SE currently do?

willvarfar 10y ago

PS: anyone want to crack an interesting problem, and earn a bounty in the process? A few days left and not a single answer on my recent question here: http://crypto.stackexchange.com/questions/37102/culling-equivalent-enigma-keys-for-short-messages http://crypto.stackexchange.com/questions/37102/culling-equi...

Strilanc 10y ago

> There should have been one SO for everything. What could possibly go right about mixing Judaism questions and Christianity questions into the programming questions?

willvarfar 10y ago
Its something I've mulled for years ;) Here's about collaborative filtering: http://williamedwardscoder.tumblr.com/post/15581427232/self-organizing-reddit http://williamedwardscoder.tumblr.com/post/15581427232/self-... Here's just general SO discontent: http://williamedwardscoder.tumblr.com/post/25426541504/stackoverflow-unwinding http://williamedwardscoder.tumblr.com/post/25426541504/stack... and http://williamedwardscoder.tumblr.com/post/26580091073/stackoverflow-humour-underflow http://williamedwardscoder.tumblr.com/post/26580091073/stack... ;)
- reitanqild 10y ago
  The collaborative filtering thing stands out as a really good idea to me. Anyone here who can say anything about the community size you would need for something like this to be viable?
reitanqild 10y ago
Ask reddit. On a more serious note: Stack Overflow started out as a technical site. Questions about programming often crosses the border to hardware or networking. They rarely however cross into theology. I'd be fine with opening a different site for theological questions as long as I could have all my technical questions in one place (and more importantly not having someone move them around.)
borispavlovic 10y ago
Is this a problem with wikipedia.org? Imagine both of these sets of questions tagged as `religion` and you have a common source for all people interested generally in any religion.
- pwdisswordfish 10y ago

reitanqild 10y ago

My thoughts as well. Wish someone would create that site. Will be happy to try to contribute. There is so much SO got correct that I think just changing a few rules could work as long as we could manage to get the word out. As managed above silos probably makes more problems than they solve (partially because of the moderators but I guess the silos should go away anyway). Same goes for strict rules about what belongs (Case in point: question about inter-company WAN traffic being rejected on networking because site guidelines says something about "inside corporate networks"). Rules about what does not belong (acceptable content) should be OK as long as they don't prevent technical questions from being asked and answered. A way to deprecate answers (and a way for viewers to "flag" them as "still useful".) And yes, a way to handle subjective questions. (Quora already has a grip of this only their IMO insane policy of sharing my every click on the site made it unusable for me and a number of others.)

CannisterFlux 10y ago

On 2 and 3 - this is dead right. As an example of 3) Here's a question that has an accepted answers that is wrong http://stackoverflow.com/questions/7635013 http://stackoverflow.com/questions/7635013 - it's regarding x86, x86_64 and x32. Since x32 was new/unknown at the time, a lot of the answers say "x32 doesn't exist". And for 2) here's an answer http://stackoverflow.com/a/26522110 http://stackoverflow.com/a/26522110 that is/was true at the time of writing - a nasty little bug in the Android NDK causes math library functions to not get linked in on pre-lollipop devices if you state you are using android platform 21 and ndk 10c. Two rules lawyers are spouting off about how the poster should post a new question or leave a comment. They are in over their heads. The comments are out of context and come across as patronizing. It is also now an example of 3) since all the tech involved is pretty old in Android terms.

tlogan 10y ago

From the point of view of person who uses SO just to find solution, incorrect answers are the biggest problem. It seems that "you don't know what you don't know" does not compute with many people involved in answering SO. But that kinda make sense: at the beginning, SO was great because answers for noob questions where hard to find. At this point, there are so many other resources to get these noob questions answered or they are already answered. Now, people start asking hard stuff - but engineers (I intentionally do not say developers) who know answers on these questions are not interested in answering.

anexprogrammer 10y ago

I used to spend quite a bit of time on SO and OnStartups. I haven't contributed to any of their sites since the ridiculous deletion of OnStartups due to the opinion of one moderator (If I remember correctly, which is vague now). I don't care that I lost half my rep. Some people had put really lengthy, researched, answers and a valuable resource was just rm'd. There's far too many silos. Sure separate maths from games from Java. You have a tech question, is it webmaster, server fault, super user, or what? Seems like there's a very slim chance of getting it right first time. Sometimes it's the remit of several but none seem to want it. Or get marked as duplicate when it's fundamentally different to the marked duplicate. Some of the subjective questions have attracted the best, or most useful, answers. Even when they've survived years they still sometimes get deleted. Stack Exchange isn't worth the effort any more.

DanielStraight 10y ago
Wow... you're not kidding about the silos. I haven't used Stack Overflow much lately, so I thought it was still just Stack Overflow and Super User. Instead... Stack Overflow Server Fault Super User Web Applications Webmasters Game Development Tex - LaTeX Unix & Linux Stack Apps Ask Ubuntu Wordpress Development User Experience Theoretical Computer Science Programmers Information Security Database Administrators Code Review Programming Puzzles & Code Golf Drupal Answers SharePoint Software Quality Assurance & Testing Computational Science Mathematica Computer Science Raspberry Pi Network Engineering Open Data Software Recommendations Arduino Joomla Data Science Emacs Vi and Vim Open Source Computer Graphics Hardware Recommendations 3D Printing Retro Computing And I'm sure I missed some. Unbelievable amount of fragmentation. Anyone want to take a crack at explaining the difference between "Computer Science" and "Theoretical Computer Science"?
- mafuy 10y ago
  I'm remotely involved there and my reply would be that TCS is research level while CS is not. There seems to be significant overlap in the questions, regardless.

danmaz74 10y ago

> make SO relevant again Curiosity: is there any more relevant alternative?

jlericson 10y ago

I sympathize with 2-4. Those things make it much harder to earn reputation by asking then it used to be. Fixing those problems requires fighting a bunch of institutional inertia, so it will require either time or some sort of judo move. But I don't understand how having separate sites relates. If you contribute to some of the smaller sites, I think you will find you'll get excellent answers from people more focused on the sort of problems the site covers. Frankly, I've seen DBA questions asked on SO that get short-sighted answers from programmers. Those questions should be asked of people who really know how a database works under the hood. I can see how collaborative filtering might help. But a big reason people have moved to dedicated sites as opposed to SO is that they want to have separate moderation policies for the questions they enjoy answering. (I've put a little bit of thought into the criteria for spinning off sites on meta.SO: http://meta.stackoverflow.com/questions/271989/does-it-pay-to-spin-off-sites http://meta.stackoverflow.com/questions/271989/does-it-pay-t...) (Disclosure: I'm a Community Manager hired by Stack Overflow.)