21
\$\begingroup\$

Stack Overflow changed the way they sort answers:

We no longer pin the accepted answer (with the green checkmark) to the top of the list of answers. By default, we now sort strictly by votes (descending order by highest score), and the accepted answer's order in the list is based on its score.

and it is asked on Meta Stack Exchange if we would like to have the accepted answer unpinned on our site:

We can change the way the engine sorts answers in site settings. We would like to hear from you all if it is something you want to see on your site. ... We are going to collect feedback before the end of September 19th.

I am posting this here so that we can have a discussion of what is best for Code Review.

\$\endgroup\$
0

4 Answers 4

Reset to default
20
\$\begingroup\$

Meh

On Code Review, it makes little difference. Code Review isn't riddled with redundant answers. Most researchers are well-prepared to read lengthy and multiple answers. The re-organization of the accepted answer's page location is rather irrelevant here.

When I am researching something on CR, I'm going to read the whole page. I'd be a fool to just read the top or accepted answer and call it quits -- what if I missed a real gem of an insight [shudders].

\$\endgroup\$
2
  • \$\begingroup\$ WE should delete this answer and put it as comments of the other two answers so that we can come to a decision. Or create a transferable vote system :-) \$\endgroup\$
    – Martin York
    Sep 30, 2021 at 4:57
  • 1
    \$\begingroup\$ That's a rather binary point of view -- one that I do not share. \$\endgroup\$
    – mickmackusa
    Sep 30, 2021 at 9:47
12
\$\begingroup\$

No

I don't think we have the same issue as Stack Overflow.

My anecdotal evidence, which is quite biased as I stay within the , is it can be good and I don't think we have the problem SO has. Having the accept pin answers has allowed late answers to hot questions be upvoted far more than they otherwise would.

Now admittedly my evidence is biased as Python only has 4 posts with a difference of 15 points or more. (Using the SEDE query in Edward's answer) And we can't really test my "can help answers rise more" observation. However we can see the post with the highest delta for Python has the following comment:

11
@holroy - yeah, I suck at code reviews where the review is basically "do it this other way" - an algorithmic review, rather than a code review. The other answers cover much of the non-algorithmic aspects also (go on, and +1 them too).
– rolfl
2015-09-14 17:31:22Z, License: CC BY-SA 3.0

I haven't seen anything bad from pinning accepts to the top. But I have seen some good.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ I agree, I think particularly self-answers have a lot more value here than SO. For example, this question codereview.stackexchange.com/q/245660 the accepted self-answer is quite late and pretty technical, but distils the other 2* higher voted answers into the key pieces of advice that OP actually used in their updated code. Often self answers represent a lot of reflection which necessarily makes them late to the party (sometimes by years!). It also helps promote the corrected version (although this can be linked to in the question). *(full disclosure; one of the answers is my own) \$\endgroup\$
    – Greedo
    Sep 9, 2021 at 14:57
11
\$\begingroup\$

YES

I think it's fine. Answers benefit the original asker, of course, but far more people read the answer than that. For that reason, leaving which answer is on top up to the community, rather than a single person seems to be more in keeping with the nature of Code Review and arguably more useful.

If you're curious, as I was, about what such questions might look like, then you can look at this SEDE query to find them.

\$\endgroup\$
4
  • 3
    \$\begingroup\$ I agree. I think SO and CR are doing it backwards. In SO the correct answer is easily identified as it is the one that solves the author's problem. Thus the accepted answer should be pinned in SO. In CR the author's code already works so the correct answer cannot be identified so easily. Thus the popularity becomes the deciding factor. In fact, I don't think we should have accepted anwers in CR at all, but that is not the topic of this discussion. \$\endgroup\$
    – TorbenPutkonen
    Sep 9, 2021 at 5:08
  • 2
    \$\begingroup\$ SO unpins the accepted answer because on SO, the primary purpose is for future people to find out the answer that works for them. That answer may not be the answer that works for the OP. So it makes sense to highlight the most popular answer over the answer that the OP happens to like. That's exactly why unpinning exists: because OPs often like less than helpful answers. Looking at it as "which answer is easiest to identify" is what's backward. Why do I care which answer helped the OP? I care which answer is mostly likely to help me. \$\endgroup\$
    – mdfst13
    Sep 9, 2021 at 8:21
  • 3
    \$\begingroup\$ I would even dare to say that we should delay the acceptance of answers for 24 to 48 hours to let more answers in and allow the community to vote. I have at least once told the OP that there is a better answer than mine and that is the one they should select. \$\endgroup\$
    – pacmaninbw
    Sep 9, 2021 at 12:13
  • \$\begingroup\$ @pacmaninbw I agree. A good percentage of users I've seen accept the first answer that's posted without giving ample time for perhaps a more experienced user to provided their input. \$\endgroup\$
    – Linny
    Sep 11, 2021 at 1:56
-3
\$\begingroup\$

YES

I think the concept of having an "accepted answer" doesn't map well on CR. What is it even supposed to mean? People here aren't looking for specific solutions, they are looking for general reviews. I think we should unpin the accepted answer and decrease its importance. We should even consider removing the notion of an accepted answer here so everyone gets a single vote.

\$\endgroup\$
2
  • 2
    \$\begingroup\$ It's commonly understood that having an accepted answer shows that the asker considers their code to be reviewed (and perhaps now working on the next iteration of the code), so removing the notion of accepted (even if it's possible) would remove a strong indication of which questions to consider answering. \$\endgroup\$
    – Toby Speight
    Sep 26, 2021 at 11:58
  • \$\begingroup\$ @TobySpeight I myself use the acceptance in a similar way, but I do believe that, it is an abuse. Why not just add a direct "Satisfied with the reviews" button? What is the point of pinning a single answer? \$\endgroup\$
    – Aykhan Hagverdili
    Sep 27, 2021 at 18:30

You must log in to answer this question.

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