4

Clean Code comments vs class documentation was closed as a duplicate of "Comments are a code smell"

There are a couple of reasons I'm finding this difficult to understand.

First, they explicitly ask about different things. The first is about documentation comments, and says:

This is not a question about in-line comments.

And the second says is about in-line comments and says:

(ie, not javadoc style method or class comments)

As a result, both questions have answers- including the accepted answer for each question- which would not be a valid answer to the other. The accepted answer to the closed question, for example, is about expressing a method's intent through it's signature, which would not have made sense on the other question.

Second, the "Comments are a code smell" question has been closed as not constructive and locked as only existing because it has "historical interest". My understanding was that the choice of which question to mark as duplicate isn't just about which is older, but which is the better question. For example, the help center describes one function of questions marked as duplicate:

but often they are left as a signpost pointing people towards the canonical answer to that question.

Given that, it seems bizarre to close a good question as a duplicate of one that would have been deleted if it weren't for "historical interest". Does it make sense to have a signpost pointing to such a question?

  • Well, that was resolved quickly! – Ben Aaronson Jun 5 '15 at 9:12
  • 2
    I agree that the questions aren't duplicates. However, that's not why I re-opened it, I re-opened it because I strongly disagree we should be closing on topic questions as duplicates of historically locked ones, even if they are duplicates. Historically locked questions are museum pieces; they are only around because they escaped deletion long enough and they aren't supposed to be used for anything. That said, if the community decides that the newer question is also too broad or primarily opinion based (the current equivalent to "not constructive"), so be it. But duplicate it is not. – yannis Jun 5 '15 at 9:20
  • @Yannis Yep, I agree, that was more or less what I was saying with my second point – Ben Aaronson Jun 5 '15 at 9:51
  • 1
    @Yannis I based my duplicate vote on this answer specifically which directly applies to the OP's question. Or this answer which hints at documentation. Or this that addresses "code smell". One could also redup to this, but as I cast a close vote already on it in the first round, I can't do it again. – user40980 Jun 5 '15 at 13:41
  • 2
    @MichaelT The question should be a duplicate to close as a duplicate, not just one of the 38 answers posted. In addition, that answer is the 10th answer down, mediocre at best, and does not address documentation comments at all. Edit: I see you've edited your comment to add more examples. I still maintain that when questions are as different as these two are, they should not be closed as duplicates. The answers there address in-line comments, while the answers to the newer question address documentation-level comments specifically and are are much better imo :) – Rachel Jun 5 '15 at 13:50
  • 2
    @MichaelT If a single answer happens to be applicable to two questions, that doesn't necessarily make the questions duplicates. What happens if somebody wants to add another answer applicable to only one of the two questions? – Ben Aaronson Jun 5 '15 at 13:53
  • @BenAaronson You are proposing that we shouldn't be able to either (A) close a question that is a target of a dup and (B) close a question as a dup of a closed question? – user40980 Jun 5 '15 at 13:56
  • @MichaelT If they're truly duplicates, then whatever close reason applied to one probably also applies to the other. So it doesn't seem to make much odds whether you close as a duplicate or for the same reason. – Ben Aaronson Jun 5 '15 at 13:58
  • @BenAaronson It is important to remember that closing as a duplicate isn't only for you. It is also for the next person to find the question and for them to have a prominent pointer to the other material on the site. – user40980 Jun 5 '15 at 14:15
  • @MichaelT I agree, I'm just not sure of the bearing that has on what we were just discussing – Ben Aaronson Jun 5 '15 at 14:22
  • @MichaelT The problem here is that a historically locked question (assuming the lock is correct) is not material we want to point people to. If it was, then it wouldn't be locked in the first place. – yannis Jun 5 '15 at 14:33
  • 1
    @Yannis if that is the case, I would encourage you to revisit the 29 links to it and create a different chain of closures. Alternatively, unlock it so that the material can be curated so that we are happy with the content and would want people to read the material there. – user40980 Jun 5 '15 at 14:35
  • @MichaelT Meh, not much to do there, there's only a handful of old questions were closed as dupes of it. – yannis Jun 5 '15 at 14:42
  • 1
    @gnat Yeah, different questions where the valid answers to one are a subset of the valid answers to the other make sense to close as duplicate. But I don't think that's the case here, I think they're two sets which happen to have a (relatively small) overlap. – Ben Aaronson Jun 5 '15 at 16:19
  • 2
    ​​​​​​​​​​​​​​​@gnat It is absolutely not okay to close two questions as duplicates if they aren't the same, whatever some random post on MSE might say. I'd go as far as saying that doing so is an abuse of the closing privilege. – yannis Jun 6 '15 at 5:38

You must log in to answer this question.

Browse other questions tagged .