This topic is my intention to bring a separate discussion -- perhaps sufficient for us to arrive at a consensus -- as to 'when' tag edits should be done. This is not the same as 'what' tags to use (which is a much bigger and more complex topic) nor how many should serve one question.

Related opinions are discussed both here and here. The first seems like the topic where this discussion might have taken place except for the fact that it did not. Two answers there relate to how many edits are acceptable and one thorough answer looks at the tags in existence and points out their view. Nevertheless, there is no point in that topic where the question of 'when' tag edits should occur is spelt out—except for the note that too many of them in close conjunction by the same is a bad idea.

However, it would be beneficial for us to spell out when exactly a tag should be removed or added to a topic.

To bring in some examples (random look at what might need a tag edit that I've been reluctant to engage in not knowing whether there would be common agreement for this). I appreciate that the "additions" are more controversial to begin with before we agree on what principles tagging should occur; nevertheless, I hope some examples to 'get the discussion going' are useful.

| |

This is and is not a counterpoint addition to semaphore's answer. I agree to that answer entirely, up to the last paragraph. The one starting with "Beyond" reads to me as perhaps a bit too much inspired by recent re-tag-edits and resulting controversies. I am unsure how much of this answer & commentary could be integrated into the existing answer, or read in minimal contrast along with the other answer? Some parts of this may read as 'disagreement' while only very few and minor parts are intended as such. Most really is 'addition':

Tag-edits as tag-only-edits/re-tag-only-edits (tag edit also means 'editing the tag/tag-info/tag-wiki):

To me, they are always acceptable — if they concern just one question that really benefits from a re-tag. A list of caveats applies, see below.

If a question benefits from improvement, then edit.
If you want to improve a question with a re-tag, then do it.

BUT: There are limits to this reasoning and thus limits to when one should edit a question in that way. Many good and correct tags to make that system work as best as possible is a quite desirable ideal, how we arrive at that in practice a road full of thorny details.

The above assumes a few priors that are not always easy to ascertain properly.

Most important of these: that the tag system itself is seen as useful, properly maintained and applied most of the time, and that usually most tags should be applied so obviously correct that there should be near instant near universal agreement on them and how they are applied to a question.

The above could be applied to systematic changes on how we improve our tag system. Like when new tags are introduced that should be applied retroactively to old questions as needed. That was done before (example), and we will likely run into this in the future. The goal should be a consistent and thus predictable way tags work and are applied. Meaning in my book: all questions should get all the tags they need as best fitting as possible, the limit is system-imposed at five tags maximum.

The caveats for the re-tag-only-edits:

  1. Do not limit yourself to re-tag-only-edit on any given post, but: A re-tag-only is always 'a minor edit' – which we should try to avoid when possible: if there is anything else that could be improved: spelling, grammar, formatting, links, references, etc. — try to improve that as well. In short: avoid tag-only-edits in favour of comprehensive edits if at all possible. As a shorthand for this spectrum of tag-edits from 'tag-only-edit' to 're-tag and renovation/comprehensive edit' I'll use 'tag-inspired-edits' (TIES) here
  2. Do limit yourself to how many TIES you want to present to the active-tab/'homepage' for any given day. 'One' seems fine to me in any case, three to five will not cause much upset, but anything above that limit will make my eyebrows go up, and some users here even get quite mad about this and see that as an abusive behaviour.
  3. This rule of thumb comes in two subspecies: 1. 'suggested edits' by low-rep users, needing review from OP, or two user reviewers or a mod review. Too much of these in a row (especially when from one single user) clog the queues and shall be avoided. 2. Hi-rep users getting carried away in 'editing frenzies': avoids the review queue, but still needs review for correctness by users. — Both casses done in excess flood the active-tab and both are undesirable. (Once I notice a pattern of the same user doing mainly minor edits, I tend to loose interest in these bumped posts altogether — that is at least for the content of that post. Instead a teacher-mode might awake and the goal is no longer reading the post as such but monitor and control those actions… or: not what I'm here for really)
  4. Do limit yourself especially when creating a new tag. It may be a very useful addition, but by now it's probably best to ask about creating it on Meta_H:SE first?
  5. Take care to look at a few special characteristics of a post:
    like: close-status and close-date: Closed questions usually need improvements (if not closed falsely/for the wrong reasons, or community disagreement). A minor edit like TIES is highly unlikely to be enough improvement to warrant a re-open. But even this minor edit will automatically put the post into the re-open-review-queue. Which in turn will lead to quite a few reviewers vote to 'leave-closed' for lack of necessary improvement. This actively sabotages the OP, the post and the reviewers of that queue.
    --> Avoid TIES / other minor edits strictly on questions closed less than 7 days ago if they aren't highly active currently anyway (which most often will imply that they are or were in that queue anyway).

(suggesting additions quite welcome)

| |

The principle downside of tag edits, as well as minor edits in general, are that they bump posts to the top and reduce visibility for new questions or new answers. When done in bulk, it floods the front page, which is all around undesirable.

That being the case, I would argue such edits are perfectly fine under these circumstances:

  1. It is a brand new question. Edit it while it's hot!

  2. The post had already been bumped to the top for whatever reason. May as well fix any issues you can see while it's up. This includes new answers, bounties, etc.

  3. You're already substantially editing a question to make genuine, concrete improvements. If so, fix the tags while you're at it.

  4. The question has a score of -5 or less. This is hidden from view, so go nuts.

Beyond this, I would say that most tag edits are so minor that it's better to hold off in general. If we focus on fixing new questions as they show up, we'll have most of the corpus sorted out sooner or later.

| |
  • 1
    Yes, this is a good point. I was half-considering suggesting there could be a "yearly clearance day" with a banner saying "look at new questions only today" where all the old stuff could be sorted. – gktscrk Jun 2 at 7:35
  • I might be worth noting again that editing a tag may inadvertently bump a closed question to the reopen cue. I noticed that happen in yesterdays tag-edit flurry. – justCal Jun 2 at 12:43
  • 1
    @gktscrk - Lol. Like the "neighborhood garage sale day" we have once a year, so residents don't have to navigate garage sale traffic all summer. – T.E.D. Jun 2 at 13:17
  • It seems that a problem of minor or tag edits is that they bump the question to the homepage. The solution shouldn't be to refrain from making those otherwise beneficial edits, but ask SE to stop bumping questions to the homepage for such minor or tag edits. – Pere Jun 3 at 11:49
  • @Pere You can do so on Meta.SE, but this isn't a new request and so far SE has not agreed. – Semaphore Jun 3 at 12:27
  • By the way, I looked at some of those -5 questions and they were at that level for a reason. I was more inclined to go for the "Delete" than "Edit tags" option at that point... – gktscrk Jun 4 at 5:38
  • 1
    @gktscrk - Thought a bit more about it, and if we ever do decide to pick an official site "touch tags day", I nominate October 8, as that's American Touch Tag Day. :-) – T.E.D. Jun 4 at 13:23

You must log in to answer this question.

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