When entering the title of this query I was pointed to here, which has tentatively endorsed this concept, especially once an answer has been accepted.

Can I get a confirmation that it is actually good practice before an answer has been accepted? (IMHO it is not.)

Two examples:

  1. My question: I didn't need a WCF answer, which I would have assumed if I had seen the question tagged with WCF even if I didn't mention it in the question or title.
  2. This question: It looks probable that the answer (if there is one) is related to AOP, but the questioner does not require it.

2 Answers 2


In general, I agree (with a little bit of personal skepticism) with this answer. The goal of the tag system, in the long run, is to allow people to find questions and answers. So as long as the new tag works towards this goal, it is a good idea. This is similar in ways to this question about Meta.

If it is a possibility of an answer... do not preemptively tag these kinds of questions. Even if the answer has been provided, don't tag it early. If a question is still seeking answers, changing the tags changes the people who are most likely to run into it, so you make it a self-fulfilling prophecy. One which the author of the question might not even want, and if you funnel the answers in the wrong direction then no one gets anything useful. If a question could be solved by, say, AOP, then wait until it is actually solved by AOP before you consider tagging it as such.

  • [The question about Meta] meta.stackexchange.com/questions/49699/… is definitely a Meta example of post-acceptance tagging, which we're agreed upon from the outcome of [the post I linked to] meta.stackexchange.com/questions/35842/… .
    – Mark Hurd
    Jun 4, 2010 at 11:10
  • 5
    tend to agree here. don't retag new questions based on possible answers; but if it's an old question, i think it's acceptable to add tags based on existing answers. in general i wouldn't replace a tag for this, only add a tag if there's room. Jun 4, 2010 at 13:02

Depends on the question. If the tag would obviously be relevant and enhance the search-friendliness of the question it may be appropriate to add such a tag. For example, once I saw a question about how to add a GUI to an existing C program. Tcl/Tk was an obvious fit to their problem from the way they described it and I answered describing how one might use Tcl/Tk to do this (Tck/Tk was originally designed to do this sort of thing).

As the original tagging on the question wasn't very good, so I retagged the question anyway and added a Tcl/Tk tag. Whether the tcl-tk tag was appripriate is something of a matter of opinion, bit I believe that it was entirely on-topic for the question.

I think this is quite a good example of a corner case where you could argue for or against the tag. I made a decision and implemented it and it probably did more good than harm to do it the way I did. Someone else could have decided differently. I don't think there's really any hard rule either way.

In your case with the WCF tag, if you didn't want to use a WCF based solution you could have clarified the question and commented on answers that suggested WCF. As the owner of the question you could have removed the tag. In the case where a tag is definitely not appropriate the owner or anyone with sufficient rep can re-tag the questoin.

So, I think it's not appropriate to make blanket rules either banning or mandating this practice. There are too many possible corner cases and the tagging is not hard and fast - it can be changed later and might signal that you need to qualify a question with (for example) a a note with a reason why a particular technology is not acceptable.

  • 2
    Just to clarify, I did remove the WCF tag and add a comment to the question -- no WCF answer had been added.
    – Mark Hurd
    Jun 4, 2010 at 11:03

You must log in to answer this question.

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