8

Many comments highlight a problem with an answer or a question. The author often acts on the comments and addresses them, making them obsolete.

It would be nice to let authors explicitly address comments as both helpful and "to be obsoleted" when updating answers/questions. (This is now typically done with yet more comments, which somehow pollute the final answer). In this sense, comments can often be seen as issue-tickets in a bug-tracking system. Once resolved, they don't add much to the discussion.

Another take on this is to provide actual context for the comment - if a comment was made while viewing version 1 of an answer, after version 3 is reached, the comment should contain a small (v1) annotation with a link to the version it was written for...

8
  • 2
    This sounds like a great idea to me!
    – Code-Guru
    Sep 11, 2012 at 23:30
  • 1
    +1 for noting the revision. I responded to a comment which, when made was was correct, however after a question was edited it had a different meaning as the words used by the commenter seemed to refer the the new code. We managed to work it out and delete our clutter comments but knowing it was made against an older revision would've avoided a pointless conversation
    – Basic
    Jul 25, 2013 at 19:55
  • 1
    What about a "mark as resolved" button on comments, which would use a voting system similar to that of flags or edits? Then, a comment could be "flagged as resolved" by either the comment's author or the community. (Edit: Never mind, looks like this is already implemented.)
    – Ajedi32
    Nov 27, 2013 at 15:18
  • I'd be happy with a "this comment refers to an older version of this post" note.
    – user152261
    Dec 21, 2013 at 16:26
  • My concern with the "to be obseleted" flag is that it's not clear how it ultimately results in the comment's deletion, and that it makes a weird allowance for using that flag and then making no edit to address the comment, which wouldn't happen if the flag was a "is now obseleted" flag, which we now/already have the "no longer needed" flag for. Nov 4, 2022 at 1:19
  • I like the idea of giving small, inline indications of which post revision a comment on the post was made at. It definitely helps make that contextual information clear when the main interface only displays the latest revision. I suggest focusing on this point for the feature request. Nov 4, 2022 at 1:20
  • Highly related to your suggestion about indicating which revision a comment was made on, there is a different feature request to display revisions inline with comments, which I also like. Your proposal is probably better at naturally preventing noise from edits that aren't related to comments, but another difference is that the other would show the revision's description inline with the comments, which has its pros. Nov 4, 2022 at 1:49
  • It would also be good to request hand-in-hand that posts show the current revision number- perhaps next to or under the edit link. Otherwise you'd still have to go to the revision list page to get that information. Nov 4, 2022 at 2:46

2 Answers 2

6

I think as a community we can do a better job of self-deleting our own comments when the issues they raise are addressed. I use comments quite often to point out flaws or limitations in an answer, or to request clarification or further details from a question. In most cases once the issue has been addressed I'll delete my comment. And in other cases where I've lost interest they may be flagged by other users.

The problem with your last paragraph is this: who is the judge that correlates the revision history with the comment's act of becoming obselete? This sounds like a lot of extra work to me, regardless of who does it.

4
  • "who is the judge that correlates the revision history with the comment's act of becoming obselete?" The reader, if its browser can inform him about revision changes. Nov 25, 2013 at 8:40
  • One of the current challenges with self-deleting is that commentors don't get notified about edits unless they follow the post, or the editor pings them with yet another comment. I suppose you could reach out in chat, but who does that? There is a feature request to notify commentors when flagging their comments. A good candidate judge for obseletion (other than the commentor) would be the editor, since they made the edit. Nov 4, 2022 at 0:48
  • @david-fong If an editor made an edit and can identify that a comment is related to that edit (and that the edit actually fully addresses the comment, which not all edits do), can't the editor just flag the comment?
    – Aaron Bertrand Staff
    Nov 4, 2022 at 0:50
  • Yes. A related problem is that there currently isn't visible guidance to the editor that that's the correct thing to do. I've been thinking and reading about this lately. Here's my feature request to add that guidance in a just-in-time fashion. This lack of guidance is what results in the current problems of slow NLN comment deletion, or "thanks I edited" reply comments. Nov 4, 2022 at 0:54
5

Your idea may be workable on slower moving questions, but I would question its efficacy on fast moving questions. Consider this flow:

OP posts question.
Reader leaves comment: "Your question suX0rs because..."
OP edits question.
OP changes their mind and edits their question again.

This flow frequently happens within the space of 5 or 10 minutes. Theoretically the question can be on to v3 or v4, but it could still be quite deficient. It is important that the original comment stays around, unversioned, during this time - then the next lot of readers can see that someone has already commented. Having the original comment versioned could mean subsequent "Your question still sucks" type comments.

Adding an option of notifying the comment author once a revision has been made is also problematic. The comment author would have to be notified for every revision during the lifetime of their comment. You then have to rely on the comment author to actually do something - I'd bet a dollar that many comments would never get actioned because comments tend to be dump-and-leave, the comment author may never come back to that question and may treat the revision notifications as just noise.

For anyone that really cares about the context of the comment, it has a timestamp which can be quite easily visually compared to the timestamp of the latest edit. If anyone is actually paying attention to the comment and cares enough they can simply flag it as obsolete.

You must log in to answer this question.

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