As part of retagging away from the Manufacturer Meta-Tag to each individual product ( and ), I've added relevant tags to questions that mentioned Mikrotik but weren't tagged as such. I'll freely admit that in a couple cases I was overzealous and added the tag where it was tangentially relevant, but there were a few cases which I still feel the edit should have been approved:

https://superuser.com/review/suggested-edits/791926 - Question about configuring Mikrotik RouterOS to serve a function; added

https://superuser.com/review/suggested-edits/791931 - Question about Mikrotik SwOS, added (Rejections were overridden by asker)

This one wasn't part of that batch, but I still feel is worth mentioning:

https://superuser.com/review/suggested-edits/760824 - Asterisk tag is rarely useful when FreePBX is being used, so I replaced with .

Why were these edits rejected? Is there something about how I approach retagging that I should change?

  • You possibly should have created a [MSU] post explaining your plans before embarking on the changes. You then reference the post in your edit description. – Burgi Sep 28 at 15:59
  • @Burgi That would make sense if the retags were rejected, but these are added tags. There was previously no Mikrotik-related tag on these questions. – Duncan X Simpson Sep 28 at 16:02
  • @Burgi Also, I did create the post; I just didn't reference it. It doesn't seem relevant for non-retags: meta.superuser.com/q/13309/341338 – Duncan X Simpson Sep 28 at 16:03
  • 1
    “I'll freely admit that in a couple cases I was overzealous and added the tag where it was tangentially relevant” - Sounds like if those were rejected, they should have been rejected, should have linked to the meta post since an edit commentary is limited. Tag has no usage guidelines, the tag should be perfect, before the entire community bumps older questions to the front page – Ramhound Sep 28 at 21:56
  • @Ramhound I'm not sure I fully understand what you're trying to say, but allow me to clarify one point: I am not contesting those instances. The only ones I disagree with are those listed. – Duncan X Simpson Sep 28 at 21:58
  • All three were 45-60 days old. None of those questions could have been improved, to encourage somebody to answer them, instead of just bumping a 2 month old question to the front page? – Ramhound Sep 28 at 22:01
  • @Ramhound I was working through the Mikrotik questions with the specific purpose of retagging them, to get rid of the Mikrotik tag. The third one seems to me to be a well-formed question, but I felt it would benefit from the freepbx tag. – Duncan X Simpson Sep 28 at 22:03
  • Retagging should be done after the community agreement, no idea how many questions you edited at once, but those three are unanswered and a tag won’t chnage that fact. But my feelings on the tag, are for that question, not this one. – Ramhound Sep 28 at 22:07
  • @Ramhound After over two weeks, there were no objections, and the tag had already been listed in the larger overall meta thread, so I decided to begin retagging. Is there something else I should have done? See link in above comment for the question. – Duncan X Simpson Sep 28 at 22:11
up vote 3 down vote accepted

There are different considerations for generally improving tagging on a question vs. cleaning up a deprecated tag.

Organized Tag Cleanup

Cleaning a deprecated tag is for the purpose of site maintenance. It ends up bumping old questions, largely with no benefit to most of the questions, but that's a side effect of the process.

The period during which a user is below 2K rep is intended to provide for learning the ropes as to what makes a good post. Edits to improve posts are reviewed to provide feedback and to avoid bad edits degrading a post. The system doesn't distinguish edits by their purpose, so <2K users are able to engage in site maintenance kinds of edits, but that's not a good use of anyone's time.

Tag cleanup efforts can be performed efficiently by users with edit privileges (2K+, and it's even simpler at 10K+ because tags can be changed without even going through the edit process). When users <2K do it, it creates a review task for two to three users who have the rep to simply do the cleanup themselves, so the cleanup requires three to four times the effort, and doesn't really contribute to the learning process.

Many reviewers don't follow the questions raised on Meta, so they may not be familiar with the tag cleanup efforts. If a <2K user engages in tag cleanup, and doesn't reference the relevant Meta question, the proposed edit is likely to be viewed in the context of a question improvement. For the majority of affected questions, this site maintenance isn't really an improvement for the question, so the edit might be rejected.

Question Improvement

Editing just to improve the tags for the benefit of the question should serve one of two purposes: help it attract answers and/or help people with a similar problem find answers.

If the existing tags are seriously problematic, we can improve either case by creating better tags, defining them with at least wiki excerpts, and retagging all of the affected questions so the tag's proper use is the rule on all of those questions rather than the exception.

New tags don't have followers. A following develops later if the tag serves it function. So tag improvement is more of an investment than an immediate benefit to those questions. No matter how good a tag is, if it is not used on most of the applicable questions, it just provides decoration without serving its intended purpose.

A tag can be critical to answering a question. For example, if the problem or solution depends on the OS or particular software or hardware, lack of a tag is substantive. Just adding that tag is likely to justify approval of an edit as long as there aren't numerous other problems that aren't fixed in the process.

But if the tag information is already covered in the question and the question is already adequately tagged, adding a supplementary tag as the only change is likely to be seen by a reviewer as not a substantive improvement. Even there, on a brand new question, if there is a chance that the additional tag might help attract answers, a reviewer might approve it, but likely wouldn't approve it as the only change to an old, inactive question. Bumping a question so that it can go back to being buried with better tags isn't really an improvement.

Another consideration is tag bloat. Many of our tags either don't serve an important tag function, or are one of many similar tags that are inconsistently used. A new tag even gets deleted by the system if at least one more question isn't tagged with it in 6 months. So if the edit consists of just a new tag, and it isn't obvious how that tag improves the question, the reviewer may see it as not a substantive improvement.

Linked Threads in the Question

The first linked thread proposed replacing the Mikrotik tag. In that question, there doesn't appear to be anything specific to Mikrotik that would affect the answer. The manufacturer tag decorated the question because that happened to be the hardware brand. Cleaning up the manufacturer tag in this case is best accomplished by removing the superfluous tag.

The second linked thread proposed adding a tag. The OP had commented "I don't think it's hardware related". So the proposed tag doesn't appear to serve an important tag function. As the only proposed change, it was not an improvement.

In the third linked thread, you proposed replacing the [asterisk] tag with [freepbx]. That would definitely not be an improvement, so rejection on that basis makes sense.

You subsequently proposed removing the [networking] tag and adding the [freepbx] tag, which was approved. That change wasn't actively harmful. It included correcting the reference to the screenshot, and perhaps the addition of [freepbx] might be useful. If the edit had been just the tag change, it might well have been considered too inconsequential; but it was combined with a correction in the body of the question.

  • Thank you for your answer! I apologize if I'm being a pain, but I'd like to fully understand what happened so I can make better edits in the future. First thread: See superuser.com/a/1351783/341338. The instructions in that answer are specific to Mikrotik - i.e. "ap bridge" is the term that RouterOS uses to describe the mode needed for this, so I believe that this was RouterOS-specific. 2. The fact that it's not hardware-related makes that tag more important, to specify the software running on said switch (SwOS). – Duncan X Simpson Oct 1 at 1:47
  • Finally, while I agree networking was the appropriate one to remove, I believe replacing Asterisk with FreePBX would have been a significant improvement. If you're familiar with cPanel, tagging that question Asterisk and not FreePBX is like tagging a question about aliases in cPanel with Exim and not cPanel. – Duncan X Simpson Oct 1 at 1:49
  • @DuncanXSimpson but in my experience is Asterisk the more commonly used, even from experts. Therfore tagging it Asterisk makes more sense, because it's easier to find. – Nordlys Jeger Oct 1 at 2:47
  • @DuncanXSimpson, in the first thread, if the solution would be different for something other than Mikrotik, then a tag could be relevant, although not necessarily required since the detail is discussed in the question and title. It's about attracting the right answerers. If we have some networking experts who know Mikrotik but not other brands, and experts who know other brands but not Mikrotik, the tag could be a useful screening tool; not so much if networking expertise generally means you are familiar with Mikrotik. (cont'd) – fixer1234 Oct 1 at 2:56
  • That's something reviewers may not know, so it would help to include the info in the edit comment. On that particular question, the tag is inconsequential, so likely to be viewed as not a real improvement, especially as the only change. Lack of the tag didn't affect getting answers, and at this point, adding it won't make a difference. (cont'd) – fixer1234 Oct 1 at 2:57
  • re: 2nd comment, FreePBX and Asterisk aren't synonymous. [freepbx] has 8 questions and 2 watchers, [asterisk] has 136 questions and 27 watchers. These tags are drawing from the same pool of answerers (likely the same people); people tend to follow the most generic version of a tag. So even if [freepbx] might be technically a better fit, [asterisk] is the better tag. It doesn't hurt to have both, but if you're going to have just one, the broader tag is better. – fixer1234 Oct 1 at 2:57
  • @fixer1234 I understand now regarding the first two; I need to give more detail in my edit comment in the future. As for freepbx vs asterisk, I didn't even think to check the watcher count. I will do so in the future. Accepting. Thank you! – Duncan X Simpson Oct 1 at 15:04

I feel that I should throw in my 2¢, because one of those suggested edit rejections was mine.

  1. I rejected it partly because there were so many other things wrong in the question that you ignored — see my subsequent edit (revision #6).  If you search the help pages and meta, you’ll see that this has been discussed before:

    • It goes without saying that we always want everything to be done well, with high quality. 
    • When you submit a suggested edit, as fixer1234 mentioned, you are making work for two or three reviewers.  I do a lot of reviews, and I try to uphold high standards.  Once a low-quality post has been brought to the attention of a reviewer like me, we are reluctant to rubber stamp it as “OK” without improving it.  If the suggested edit fixes more than 50% of the problems that were there (and doesn’t introduce any new problems!), I’m generally willing to click on “Improve”, which means I approve the suggested edit and then go on immediately to do an additional edit of my own.  But otherwise (especially if the edit actually causes damage; this was not the case with yours), I feel that the suggested edit is making more work for the reviewer than the editor did themselves.  In such a case, I often “Reject and Edit”, which is fairly self-explanatory — I reject the suggested edit and then go on immediately to do an alternative edit of my own.  This may include fixing some of the same problems that the suggested edit tried to fix.
  2. You neglected to mention (but fixer1234 noticed) that you resubmitted largely the same suggested edit later that day, and I approved that one.

    • Your edit comment (on the edit that I rejected) said “Replaced asterisk tag with freepbx”.  The question mentioned “FreePBX” twice and “asterisk” eight times (including the existing tag, and four log messages).  Why should anybody approve such an edit?  It’s like replacing the engine of a car with a microwave oven — it looked like you were replacing a more important component with a less important one, with no explanation.
    • You say “Asterisk tag is rarely useful when FreePBX is being used …”  Well, I don’t know that.  Perhaps you should

      • create a meta post making that claim (and supporting it with an objective argument),
      • wait for consensus from the community, and
      • edit the and tag wikis to reflect that.

      Currently the tag wiki is more than 3.3 times larger than the tag wiki.  There’s no reason for anybody to believe or even suspect that the FreePBX tag is in any way preferable to the Asterisk tag.

      And I see now that you and fixer1234 already discussed this in comments here.

    • I seem to recall that you and I exchanged some comments between your two suggested edits.  I don’t see them now, and I don’t have any record of what we said.
    • The comment of your second edit, “Changed FreePBX to Asterisk (that's a FreePBX screenshot); removed networking tag and added freepbx”,
      • explained a little about the rationale for adding the freepbx tag, and
      • added the freepbx tag, retaining the asterisk tag, by deleting the networking tag.  I seem to recall that I suggested that, because the networking tag was so general, and the question already had the and tags.

    Bottom line: not all improvements are obviously improvements.  It’s often useful to explain why a change is an improvement — especially when you’re suggesting an edit.

  • Thanks for the explanation. I neglected to mention that, as I had forgotten about it. I used SEDE to see my rejected edits, in an attempt to learn why certain edits were rejected and improve my editing overall. The consequence of that was I lacked that context. If you're interested in why I said that about asterisk not being useful there, think of it as similar to a web hosting control panel in this sense: meta.serverfault.com/a/9405/295551 – Duncan X Simpson Oct 10 at 19:11

You must log in to answer this question.

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