As an example of 3) Here's a question that has an accepted answers that is wrong 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 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.
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.
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"?(disclosure: 11K rep, mostly on tools like make, autofoo and cmake.)
But maybe this would lead to it's own problems too.
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-t...)
(Disclosure: I'm a Community Manager hired by Stack Overflow.)
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.
What could possibly go right about mixing Judaism questions and Christianity questions into the programming questions?
Here's about collaborative filtering: http://williamedwardscoder.tumblr.com/post/15581427232/self-...
Here's just general SO discontent: http://williamedwardscoder.tumblr.com/post/25426541504/stack... and http://williamedwardscoder.tumblr.com/post/26580091073/stack...
;)
Anyone here who can say anything about the community size you would need for something like this to be viable?
https://en.wikipedia.org/wiki/Wikipedia:Lamest_edit_wars
https://en.wikipedia.org/wiki/Wikipedia:Arbitration/Index/Ca...
Also, there's a very good reason "the dramaboards" are called that.
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.)
And the fear of computers into God, perhaps.
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.)
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-equi...
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)