9 ms·

Meh. Stack Overflow is suffering from the same "problem" as Wikipedia: the quality & quantity of existing content is now so high that it's becoming really hard

Meh. Stack Overflow is suffering from the same "problem" as Wikipedia: the quality & quantity of existing content is now so high that it's becoming really hard to contribute new content. However, while this is indeed a real problem for "askers" trying to get heard in all the noise, and there's diminishing returns for the "answerer" who has less fun questions to tackle, it's also completely irrelevant to the 99% of us whose questions have already been asked & answered, and those answers can be pulled up in seconds by your favorite search engine. Also, in case the blog author's name (Jon Skeet) didn't ring a bell, he's Stack Overflow's #1 contributor. If you've ever searched for anything related to C#, you've probably seen his answers, which have racked up over a million reputation points: https://stackoverflow.com/users/22656/jon-skeet https://stackoverflow.com/users/22656/jon-skeet

mooreds 9y ago
This is really insightful. In some ways every community goes through the "wild west" -> settlers -> cities evolution. Happened with the internet at large, happens with most technologies that achieve mass adoption, and it appears to have happened with Wikipedia and Stackoverflow. (I read an interesting article about technology pioneers, settlers and other archetypes, but I couldn't find it.) Once most of the pioneering work is done, and the low hanging articles are written, a community needs to decide how to proceed. Who are you going to maximize value for? In the early days it makes sense to focus on what is the rarest resource (in SO's case, probably the answerers) but as a community matures, the focus can shift. Now it is frankly the anonymous viewer that it makes sense to optimize for, both from a monetary standpoint (jobs display being a primary monetization strategy) and a social good standpoint (far more anonymous viewers than either askers or answerers).
- mooreds 9y ago
  Here's the article (someone else found it for me!): http://blog.gardeviance.org/2015/03/on-pioneers-settlers-town-planners-and.html http://blog.gardeviance.org/2015/03/on-pioneers-settlers-tow...
zombieprocesses 9y ago
> Also, in case the blog author's name (Jon Skeet) didn't ring a bell The guy is a legend. If there was ever an MVP in the C#/.NET space, it's this guy. His books are great too.
- acjohnson55 9y ago
  And also author of one of the more epic StackOverflow answers: https://stackoverflow.com/a/6841479/807674 https://stackoverflow.com/a/6841479/807674
- leetcrew 9y ago
  > If there was ever an MVP in the C#/.NET space, it's this guy. he has (had?) some competition from eric lippert. i used to love reading his blog when i did c# work.
psyc 9y ago
I was going to say the same exact thing, that it has the same problems as wikipedia. But then I was going to say "hard to contribute ... because control-freak insiders who've managed to take political control of what should have been an Internet-wide mass-community don't want you to."
- isaachier 9y ago
  Not sure what is political. I consider Wikipedia and SO to be perfect democracies of information. People upvote answers they consider correct. Those who contribute good answers frequently have demonstrated their skills and knowledge. If another talented contributor provides intelligent answers, we would happily upvote his or her answers as well.
  - eat_veggies 9y ago
    That would be the case if moderators couldn't arbitrarily close questions or revert edits on a whim.
- gbacon 9y ago
  Sounds like another case of geeks versus wonks. In Tucker’s “A Political Theory of Geeks and Wonks,” he delineates the latter with Political wonks are fascinated by process. They love the game. They get as much satisfaction from observing as changing. They want to be players above all else. Ideals bore them. History is mere data. Intellectuals seem irrelevant. What matters to the wonk are the hard realities of the ongoing political struggle. They defer to title and rank. They thrive on meetings, small victories, administrative details, and gossip about these matters. Knowing who is who and what is what is the very pith of life. Geeks, on the other hand: They are no less fascinated by detail but are drawn to ideals. Observation alone bores them. They are drawn to the prospect of change. They don't want to be players as such; they question the very rules of the game and want to change them. They are happy to make a difference in the ideological infrastructure, whether big or small. They tend to work alone and totally disregard caste distinctions. As to your frustration with Wikipedia’s political elite The wonks are the ones who consolidate, stabilize, and entrench the status quo; the geeks are the ones who prepare revolutionary change. The wonks freeze it into place and make it work more efficiently; the geeks imagine and work toward a future that no one thought was possible. The wonks rule out drastic and extreme measures as imprudent and reckless; they geeks think these paths are the only ones worth pursing, and have confidence that the unknown future will somehow work itself out. The wonks try to bring the king around to their point of view; the geeks kill the king. Tucker describes the conflict between the two camps. The geeks and wonks can work together but there will always be a natural tensions between the two. The wonks think the geeks are hopeless, powerless, reckless outsiders whose heads are full of useless and unrealistic fantasies. The geeks think that the wonks are part of the system and, therefore, more than likely corrupted by it, and increasingly so. Broadening the view, the struggle to control history is a battle between the wonks and the geeks … According to his view, Julius Caesar represented the wonks and Brutus the geeks in Ancient Rome. In the America’s revolutionary generation, they were Alexander Hamilton and Thomas Jefferson.
jasode 9y ago
>suffering from the same "problem" as Wikipedia: the quality & quantity of existing content is now so high that it's becoming really hard to contribute new content. However, while this is indeed a real problem for "askers" trying to get heard in all the noise, and there's diminishing returns for the "answerer" who has less fun questions to tackle, I agree with your observation but Jon Skeet's essay seemed to be focusing on something else... the culture of hostility (or perceived hostility). The key words I noticed he used (multiple times) were "hostility", "jerks", "feelings". On the other hand, the keywords I would associate with your observation would be... "boredom", "novelty of a new website", "low hanging fruit to quickly build easy karma", etc. To discuss Jon's specific angle, let's use an example of a [tag] that's fairly new technology such as "[vue.js]". Since Vue is 2014, it doesn't have the "completeness of answers" of a very old tag like C# from 2008. Here's a Vue question sitting downvoted to negative 4: https://stackoverflow.com/questions/49345098/how-to-get-props-in-vue https://stackoverflow.com/questions/49345098/how-to-get-prop... Maybe his question deserved all those downvotes. I don't know enough about Vue to guess if the downvotes were all caused by a pasted screenshot instead of pure text. To Jon's "culture" essay, is the minus-4 the optimal way to communicate it's a "bad" question? Is it a bad question because of its format (the screenshot) or was it bad because "props" is a trivial LMGTFY (Let Me Google That For You Question)? Here's another Vue question at negative 1: https://stackoverflow.com/questions/49321665/how-to-add-a-get-parameter-to-a-file-obtained-by-lazy-loading-of-vue-router https://stackoverflow.com/questions/49321665/how-to-add-a-ge... To generalize, if there's "[new technology]" that opens a vast untaken area for new contributions of answers, is the culture created by Stackoverflow's mechanisms (the downvotes, the karma system, the moderators' powers, the "closings" vs "moving", etc) ... the best incentive system to minimize hostility? Are Jon's well-intentioned "covenants" realistic? Or are they naive to human tendencies when any website is open to mass participation?
- mark-r 9y ago
  That's one thing Jon didn't touch on, the downvote. It's the easiest thing in the world for the giver, but to the receiver it feels like a kick in the gut. It definitely contributes to the feeling of hostility. Personally I try to use my downvotes only as a last resort, for factually wrong answers or questions where clarification attempts went nowhere.
klenwell 9y ago
> this is indeed a real problem for "askers" trying to get heard in all the noise, and there's diminishing returns for the "answerer" who has less fun questions to tackle To address this issue, I put together a "deep-cuts" algorithm that uses the API to tries to find overlooked higher quality questions that have yet to be answered. The code is available here: https://github.com/klenwell/stack-overflow-deep-cuts/blob/master/python/models/scored_question.py https://github.com/klenwell/stack-overflow-deep-cuts/blob/ma... It's not perfect. But if you're looking to engage with Stack Overflow, I've found it much more useful than the rat-race to answer (or close) new questions.
- tomca32 9y ago
  This looks really useful. I'm surprised SO doesn't already have that feature.
  - tneu89n43nth 9y ago
    Actually, they do. It's less obvious with the latest redesign, but if you click on "Questions" next to the logo in the header, it takes you to the list of questions that have been asked. At the top of that list is a set of tabs, and the last one is "unanswered." If you click on it, you see the highest ranked, not yet answered questions. I'm not sure if there's anything more to it than just the rank, but it's a good place to start.
minimaxir 9y ago
You'd think that there'd be a low Q/A rate, but apparently it depends significantly by tag, with some tags having low rates (10-20%: http://minimaxir.com/img/stack-overflow-tags/acceptable_answer_bottom_30.png http://minimaxir.com/img/stack-overflow-tags/acceptable_answ...) and some having high rates (60%-70%: http://minimaxir.com/img/stack-overflow-tags/acceptable_answer_top_30.png http://minimaxir.com/img/stack-overflow-tags/acceptable_answ...) And some pairs do better/worse than others: http://minimaxir.com/img/stack-overflow-tags/so_tag_adjacency_percent.png http://minimaxir.com/img/stack-overflow-tags/so_tag_adjacenc...
edem 9y ago
I even bumped into him with some Java questions.

mooreds 9y ago

This is really insightful. In some ways every community goes through the "wild west" -> settlers -> cities evolution. Happened with the internet at large, happens with most technologies that achieve mass adoption, and it appears to have happened with Wikipedia and Stackoverflow. (I read an interesting article about technology pioneers, settlers and other archetypes, but I couldn't find it.) Once most of the pioneering work is done, and the low hanging articles are written, a community needs to decide how to proceed. Who are you going to maximize value for? In the early days it makes sense to focus on what is the rarest resource (in SO's case, probably the answerers) but as a community matures, the focus can shift. Now it is frankly the anonymous viewer that it makes sense to optimize for, both from a monetary standpoint (jobs display being a primary monetization strategy) and a social good standpoint (far more anonymous viewers than either askers or answerers).

mooreds 9y ago
Here's the article (someone else found it for me!): http://blog.gardeviance.org/2015/03/on-pioneers-settlers-town-planners-and.html http://blog.gardeviance.org/2015/03/on-pioneers-settlers-tow...

zombieprocesses 9y ago

> Also, in case the blog author's name (Jon Skeet) didn't ring a bell The guy is a legend. If there was ever an MVP in the C#/.NET space, it's this guy. His books are great too.

acjohnson55 9y ago
And also author of one of the more epic StackOverflow answers: https://stackoverflow.com/a/6841479/807674 https://stackoverflow.com/a/6841479/807674
leetcrew 9y ago
> If there was ever an MVP in the C#/.NET space, it's this guy. he has (had?) some competition from eric lippert. i used to love reading his blog when i did c# work.

psyc 9y ago

I was going to say the same exact thing, that it has the same problems as wikipedia. But then I was going to say "hard to contribute ... because control-freak insiders who've managed to take political control of what should have been an Internet-wide mass-community don't want you to."

isaachier 9y ago
Not sure what is political. I consider Wikipedia and SO to be perfect democracies of information. People upvote answers they consider correct. Those who contribute good answers frequently have demonstrated their skills and knowledge. If another talented contributor provides intelligent answers, we would happily upvote his or her answers as well.
- eat_veggies 9y ago
  That would be the case if moderators couldn't arbitrarily close questions or revert edits on a whim.
gbacon 9y ago
Sounds like another case of geeks versus wonks. In Tucker’s “A Political Theory of Geeks and Wonks,” he delineates the latter with Political wonks are fascinated by process. They love the game. They get as much satisfaction from observing as changing. They want to be players above all else. Ideals bore them. History is mere data. Intellectuals seem irrelevant. What matters to the wonk are the hard realities of the ongoing political struggle. They defer to title and rank. They thrive on meetings, small victories, administrative details, and gossip about these matters. Knowing who is who and what is what is the very pith of life. Geeks, on the other hand: They are no less fascinated by detail but are drawn to ideals. Observation alone bores them. They are drawn to the prospect of change. They don't want to be players as such; they question the very rules of the game and want to change them. They are happy to make a difference in the ideological infrastructure, whether big or small. They tend to work alone and totally disregard caste distinctions. As to your frustration with Wikipedia’s political elite The wonks are the ones who consolidate, stabilize, and entrench the status quo; the geeks are the ones who prepare revolutionary change. The wonks freeze it into place and make it work more efficiently; the geeks imagine and work toward a future that no one thought was possible. The wonks rule out drastic and extreme measures as imprudent and reckless; they geeks think these paths are the only ones worth pursing, and have confidence that the unknown future will somehow work itself out. The wonks try to bring the king around to their point of view; the geeks kill the king. Tucker describes the conflict between the two camps. The geeks and wonks can work together but there will always be a natural tensions between the two. The wonks think the geeks are hopeless, powerless, reckless outsiders whose heads are full of useless and unrealistic fantasies. The geeks think that the wonks are part of the system and, therefore, more than likely corrupted by it, and increasingly so. Broadening the view, the struggle to control history is a battle between the wonks and the geeks … According to his view, Julius Caesar represented the wonks and Brutus the geeks in Ancient Rome. In the America’s revolutionary generation, they were Alexander Hamilton and Thomas Jefferson.

jasode 9y ago

>suffering from the same "problem" as Wikipedia: the quality & quantity of existing content is now so high that it's becoming really hard to contribute new content. However, while this is indeed a real problem for "askers" trying to get heard in all the noise, and there's diminishing returns for the "answerer" who has less fun questions to tackle, I agree with your observation but Jon Skeet's essay seemed to be focusing on something else... the culture of hostility (or perceived hostility). The key words I noticed he used (multiple times) were "hostility", "jerks", "feelings". On the other hand, the keywords I would associate with your observation would be... "boredom", "novelty of a new website", "low hanging fruit to quickly build easy karma", etc. To discuss Jon's specific angle, let's use an example of a [tag] that's fairly new technology such as "[vue.js]". Since Vue is 2014, it doesn't have the "completeness of answers" of a very old tag like C# from 2008. Here's a Vue question sitting downvoted to negative 4: https://stackoverflow.com/questions/49345098/how-to-get-props-in-vue https://stackoverflow.com/questions/49345098/how-to-get-prop... Maybe his question deserved all those downvotes. I don't know enough about Vue to guess if the downvotes were all caused by a pasted screenshot instead of pure text. To Jon's "culture" essay, is the minus-4 the optimal way to communicate it's a "bad" question? Is it a bad question because of its format (the screenshot) or was it bad because "props" is a trivial LMGTFY (Let Me Google That For You Question)? Here's another Vue question at negative 1: https://stackoverflow.com/questions/49321665/how-to-add-a-get-parameter-to-a-file-obtained-by-lazy-loading-of-vue-router https://stackoverflow.com/questions/49321665/how-to-add-a-ge... To generalize, if there's "[new technology]" that opens a vast untaken area for new contributions of answers, is the culture created by Stackoverflow's mechanisms (the downvotes, the karma system, the moderators' powers, the "closings" vs "moving", etc) ... the best incentive system to minimize hostility? Are Jon's well-intentioned "covenants" realistic? Or are they naive to human tendencies when any website is open to mass participation?

mark-r 9y ago
That's one thing Jon didn't touch on, the downvote. It's the easiest thing in the world for the giver, but to the receiver it feels like a kick in the gut. It definitely contributes to the feeling of hostility. Personally I try to use my downvotes only as a last resort, for factually wrong answers or questions where clarification attempts went nowhere.

klenwell 9y ago

> this is indeed a real problem for "askers" trying to get heard in all the noise, and there's diminishing returns for the "answerer" who has less fun questions to tackle To address this issue, I put together a "deep-cuts" algorithm that uses the API to tries to find overlooked higher quality questions that have yet to be answered. The code is available here: https://github.com/klenwell/stack-overflow-deep-cuts/blob/master/python/models/scored_question.py https://github.com/klenwell/stack-overflow-deep-cuts/blob/ma... It's not perfect. But if you're looking to engage with Stack Overflow, I've found it much more useful than the rat-race to answer (or close) new questions.

tomca32 9y ago
This looks really useful. I'm surprised SO doesn't already have that feature.
- tneu89n43nth 9y ago
  Actually, they do. It's less obvious with the latest redesign, but if you click on "Questions" next to the logo in the header, it takes you to the list of questions that have been asked. At the top of that list is a set of tabs, and the last one is "unanswered." If you click on it, you see the highest ranked, not yet answered questions. I'm not sure if there's anything more to it than just the rank, but it's a good place to start.

minimaxir 9y ago

You'd think that there'd be a low Q/A rate, but apparently it depends significantly by tag, with some tags having low rates (10-20%: http://minimaxir.com/img/stack-overflow-tags/acceptable_answer_bottom_30.png http://minimaxir.com/img/stack-overflow-tags/acceptable_answ...) and some having high rates (60%-70%: http://minimaxir.com/img/stack-overflow-tags/acceptable_answer_top_30.png http://minimaxir.com/img/stack-overflow-tags/acceptable_answ...) And some pairs do better/worse than others: http://minimaxir.com/img/stack-overflow-tags/so_tag_adjacency_percent.png http://minimaxir.com/img/stack-overflow-tags/so_tag_adjacenc...

edem 9y ago

I even bumped into him with some Java questions.