3 ms·

Judging by the number of answers I see which were valid and good answers in 2011 and are horrendously outdated now, it's not just grown up, it's grown long in t

Judging by the number of answers I see which were valid and good answers in 2011 and are horrendously outdated now, it's not just grown up, it's grown long in the tooth.

dudul 8y ago
That's a really good observation. A lot of very highly ranked answers no longer apply, and still they are very visible due to their score. If the idea is to be the "wikipedia of programming", I wonder if "votes" are enough. Should there be an active initiative to prune stupid/bad questions? Revisit old answers?
- Flimm 8y ago
  It's actually worse than that. The accepted answer by the OP is pinned to the top forever (or until OP changes his or her mind, which is often never), no matter how many downvotes the pinned answer collects.
  - superflyguy 8y ago
    Possibly just need to reduce every answer's score a little every month to favour answers which are actively being voted up more recently.
- sureaboutthis 8y ago
  Then you vote to close the question. It may take a while but, eventually, it will be deleted. Just yesterday, I closed three or four such questions in a brief visit.
umvi 8y ago
The only questions I've seen constantly outdated are web dev related. More stable technologies like C/C++/python seem to have up to date answers most of the time (purely anecdotal).
- cdoxsey 8y ago
  Lately I've been working with c++ and, at least in my experience, as a developer who's native tongue isn't c++, many of the answers have failed to keep up with the radical changes in the language in the last 10 years. Post-17 C++ is completely different from pre-11 C++. That said there's usually an answer further down the page with a more up to date solution, it's just not the accepted or most popular one.
  - umvi 8y ago
    True, but consider your user base. I, for one, maintain legacy C/C++98 code. C++17 answers are not helpful to me because the legacy system does not have a new enough compiler, sysroot, etc for me to take advantage of them. It's not like web development where users' browsers auto-update and you can start immediately using the newest versions of everything. Only our newest projects are using modern C++ (14). I would wager the majority of C++ users "in the field" are using 11 and below and haven't had a good opportunity to switch to 14 or 17 yet. As such, answers that seem outdated to you may be a good fit for the majority. It's sort of like how any native answers automatically float higher than boost answers - why is this? Because most c++ users are not using boost and have no interest in using boost.
yAnonymous 8y ago
Try to update one of those outdated answers and your changes will be removed, because the original author has more internet points. The amount of circle jerking on Stack Overflow is stunning.

Judging by the number of answers I see which were valid and good answers in 2011 and are horrendously outdated now, it's not just grown up, it's grown long in the tooth.

dudul 8y ago

That's a really good observation. A lot of very highly ranked answers no longer apply, and still they are very visible due to their score. If the idea is to be the "wikipedia of programming", I wonder if "votes" are enough. Should there be an active initiative to prune stupid/bad questions? Revisit old answers?

Flimm 8y ago
It's actually worse than that. The accepted answer by the OP is pinned to the top forever (or until OP changes his or her mind, which is often never), no matter how many downvotes the pinned answer collects.
- superflyguy 8y ago
  Possibly just need to reduce every answer's score a little every month to favour answers which are actively being voted up more recently.
sureaboutthis 8y ago
Then you vote to close the question. It may take a while but, eventually, it will be deleted. Just yesterday, I closed three or four such questions in a brief visit.

umvi 8y ago

The only questions I've seen constantly outdated are web dev related. More stable technologies like C/C++/python seem to have up to date answers most of the time (purely anecdotal).

cdoxsey 8y ago
Lately I've been working with c++ and, at least in my experience, as a developer who's native tongue isn't c++, many of the answers have failed to keep up with the radical changes in the language in the last 10 years. Post-17 C++ is completely different from pre-11 C++. That said there's usually an answer further down the page with a more up to date solution, it's just not the accepted or most popular one.
- umvi 8y ago
  True, but consider your user base. I, for one, maintain legacy C/C++98 code. C++17 answers are not helpful to me because the legacy system does not have a new enough compiler, sysroot, etc for me to take advantage of them. It's not like web development where users' browsers auto-update and you can start immediately using the newest versions of everything. Only our newest projects are using modern C++ (14). I would wager the majority of C++ users "in the field" are using 11 and below and haven't had a good opportunity to switch to 14 or 17 yet. As such, answers that seem outdated to you may be a good fit for the majority. It's sort of like how any native answers automatically float higher than boost answers - why is this? Because most c++ users are not using boost and have no interest in using boost.

yAnonymous 8y ago

Try to update one of those outdated answers and your changes will be removed, because the original author has more internet points. The amount of circle jerking on Stack Overflow is stunning.