19
\$\begingroup\$

Today, Stack Exchange changed the way answers are sorted on Stack Overflow, meaning that accepting an answer no longer pins it to the top (regardless of sorting method). It still has the green checkmark next to it, and accepting an answer is still exactly the same for the purposes of reputation, badges, green marker next to a question etc.

Stack Exchange are also asking the other sites in the network whether they'd like this behaviour or not, as this is now a site-configurable setting.

We've made noise before about removing or modifying the accept feature, as it doesn't exactly fit as intended with our site's goals, especially the part about a human accepting an answer that may be the wrong answer (or that may become outdated in the future). This isn't quite removing the accept feature, but it's a similar, milder step.

Therefore, should we request that accepted answer be unpinned on this site, or should we keep accepted answers pinned to the top?

\$\endgroup\$

3 Answers 3

25
\$\begingroup\$

Yes! Emphatically yes.

The accepted answer is rarely the most important answer. In part because often the accepted answer is out of date and thus not really even the best scoring, but more importantly because being the best scoring overall does not mean it is the cleverest, or most interesting. For code-golf it just means its the shortest answer from the handful of good golfing languages, which is not on it's own interesting.

\$\endgroup\$
6
  • 3
    \$\begingroup\$ The problem is that for challenges like king-of-the-hill, being able to pin answers is useful. It's not great for old code golf challenges with accepted answers, but aside from a few those don't get much traffic anyway from users new enough to the site to not know how to vote. \$\endgroup\$ Commented Sep 9, 2021 at 16:51
  • 2
    \$\begingroup\$ @BrowncatPrograms I'm not sure how usefully it actually is if you have a score board in the challenge. And a scoreboard is much more useful for a variety of purposes: It can be maintained by the community instead of falling on one person; It shows second and third place instead of just first; it doesn't interfere with user selected sorting methods etc. \$\endgroup\$
    – Wheat Wizard Mod
    Commented Sep 9, 2021 at 17:02
  • \$\begingroup\$ Well sure, a scoreboard also indicates who's in the lead, but it doesn't fulfill the same role as pinning the answer (it's sort of like the green checkmark on its own). Having the winning answer at the top is desirable for KotHs, as they often do deserve the upvotes, and usually take longer to develop (so they miss out on the initial wave of upvotes on simpler bots, and might even be on the second page for large KotHs). \$\endgroup\$ Commented Sep 9, 2021 at 18:06
  • 1
    \$\begingroup\$ @BrowncatPrograms We are not going to fix the issues with voting by abusing a feature designed with someone else entirely in mind. At the same time it's still very annoying even on Koths because it breaks the sorting. \$\endgroup\$
    – Wheat Wizard Mod
    Commented Sep 9, 2021 at 18:53
  • 1
    \$\begingroup\$ The whole point of pinning is that it breaks the sorting, right? And even if the feature wasn't intended for pinning the winners of KotHs, it is intended for pinning the best solution to a problem on someplace like SO, and a KotH does have an unambiguously best solution (which is much more often correlated to how interesting it is, unlike in code golf). The benefit of helping with voting issues isn't the main reason, but it is a great advantage of having them pinned. I don't see how that's a disadvantage or "abusing a feature". \$\endgroup\$ Commented Sep 11, 2021 at 15:26
  • 2
    \$\begingroup\$ I love that this answer is pinned. Irony, sweet irony. \$\endgroup\$
    – AviFS
    Commented Sep 13, 2021 at 8:29
10
\$\begingroup\$

Accepted answers should not be pinned

As mentioned in chat, it can be really nice to have the accepted answer pinned can be nice for things like KotHs or questions with a definite winner, however I've personally seen accepted answers used much more often in things like challenges where there probably shouldn't be an accepted answer, whether it's from a new user or from an older challenge.

In a perfect world, the best thing would be for accepted answers to be pinned, and for us to only accept answers on challenges/questions that where it makes sense, such as KotHs, or questions on how to golf a specific piece of code. However, this isn't really feasible, since there are already so many challenges with accepted answers, as well as new users who don't know better.

In my opinion, given our circumstances, having accepted answers unpinned would be better for us overall, since the majority of our site and our policies are geared towards challenges where there is no "right answer," such as questions, or challenges where there is no one winner, since each language competes against its own type of language.

\$\endgroup\$
4
\$\begingroup\$

Accepted answers should be pinned

We rarely accept answers. Sometimes when an answer is accepted it's because it's on an old question and it was the shortest, but it's quite often used in challenges like KotHs to indicate which answer is winning.

The ability to accept an answer to pin it is useful for some types of challenges, and not overly harmful when misused. It's not great for old code golf challenges with accepted answers, but aside from a few those don't get much traffic anyway from users new enough to the site to not know how to vote.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ I'd be for this if we could restrict accepting answers to specific kinds of challenges, but as is many accepted answers are plain code golfs. \$\endgroup\$
    – xnor
    Commented Sep 10, 2021 at 23:03
  • 4
    \$\begingroup\$ I really don't think it's such a problem having accepted answers on plain code golfs pinned, and it is such a problem to lose them on KotH/CnR/etc \$\endgroup\$
    – pxeger
    Commented Sep 11, 2021 at 13:32

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .