Should answers refrain from referring to other answers on the same question?

Question

Reviews don't pile up at the same time. Whereas Alice was already awake when Charlie's question was posted and wrote an excellent review first, Bob took a longer nap, had a coffee, and then noticed that there is another variant that also reviews the question. He begins his post with the following paragraph:

Alice already points out where your Widgets have defects, so let's focus on your Gadgets.

Bob's answer is also excellent. He points out the strengths and weaknesses of Charlie's Gadgets in the rest of his post.

---

This scenario happens often, although with different levels of quality. Occasionally we end up with answers that start with "as @user said , "continuing from @user" or "@user [...] answer [..] I'll just add the following".

Most of those remarks can be left out without changing the significance of the answer. After all, we're allowed to only look at parts of the code. We don't need to justify that we only covered Gadgets. Unless Bob refers to the content of Alice' answer directly,* it's just additional noise. And while it's unlikely on CR, Alice might delete her answer later and Eve doesn't understand the reference in Bobs answer.

Should we refrain from referring to other answers if we don't actually use any content from those answers?

If the answer is "yes", how should we deal with the situation where one refers to another answer?

• Comment?
• Edit?
• Comment and edit?

^{* that would need proper attribution and is a whole other beast}

Answer 1

I regularly reference other people's answers when I answer questions. I typically also include a link to that answer. I feel it is important to show where the answers are different, disagree, or complement my own answer, or even why my answer does not cover a certain aspect of the code because another answer does a good job of it already.

it is my hope that people will click the link, read the other answers, and agree with me, and vote for those answers too.

On stack overflow it's often the case that people vote for only 1 answer they "agree with", but on Code Review I want people to vote as many times as possible on all answers that have a positive contribution. Complimenting another answer, helping voters find that answer, and differentiating my answer from theirs is important, I feel.

For examples of where I have done this before:

• Finding pair of sum in sorted array
• Using synchronized blocks and volatile variables to ask a Java thread to terminate gracefully
• Knight moves in Chess game
• ....

While I feel it is a positive thing to do, I also now recognize that there may be a down-side to it too. Still thinking about whether it's a good practice in the long term.

Answer 2

I don't think we should limit people from this type of "teamwork", unless it is rude or "not nice" or destructive.

Someone may see something that should be followed in conjunction with another answer, but they should not edit the other answer.

Sometimes a comment on someone's answer about gotchas can turn into a larger beast than can be contained in a single comment or without sufficient code blocks.

Your first example is a reference to your comment, but it makes a valid point so that the OP doesn't push back on the answer, saying that their question wasn't answered.

I say this because in the original version of the question the user was asking specifically why the examiner was dissatisfied with the code, this user reiterated what you said because it is something that they would have said anyway as a way of pushing that issue aside to actually review the code itself, without trying to figure out why the examiner was dissatisfied(something we cannot answer).

in your third example you had said in your answer that

I know almost no F#. But I know Haskell. So I think I should be able to give some insight. An F# expert can give you more, of course

The user was simply stating that you gave a good review, they were backing you up as an "expert" in F# so to speak, giving your answer confidence.

I would be more concerned with the "disclaimer" at the top of your answer, but I can understand worrying about syntax errors.

tl;dr I wouldn't worry about someone referencing your answers in the way that you have shown in your examples, I would be impressed or honored (maybe not the right words) because they are telling you and the OP that you right and that they should up-vote your answer

Answer 3

I agree with @rolfl about linking/mentioning other answers and I find that the benefits of doing that outweigh by far any downsides.

Code Review answers usually do not compete with each other but rather build a big picture. Bringing a little bit of structure to that picture helps to better understand it. This also shows that people actually read other answers in order to provide more information and create an even sharper picture.

Forbiding linking other answer would only harm the quality of CR.

I sometimes even add links to answers that were posted after mine because I found they were very good and didn't get much attention yet.

I'm always happy when someone links/mentions my answer because this is also some kind of a recognition. Votes are anonymous, links are not. And if they criticize your answer, then you know it was yours so you can also learn something from it.

It be a great feature if SE could notify you when your answer is linked/mentiond anywhere or list all links to other posts that link yours.

Answer 4

I can see why you say that it's noise, but I think that to say that

[i]t's just noise

is to go too far. It's context. While you're correct that

We don't need to justify that we only covered Gadgets.

(my emphasis), there is value in distinguishing between the possibilities that

1. I only looked at the code for Gadgets, so there's only been one pair of eyes on the Widgets;
2. I looked at the Widgets but I didn't see anything which Alice had missed.