I have basically this question. But that 2012 post doesn't work for multiple reasons:

• The OP has forgotten to add good reasons legitimising his request. That lead to the usual avalanche of downvotes and "you don't want that" comments
• The accepted, upvoted and only answer is half rejecting the question altogether, and half giving a very unsatisfactory answer (replace a core library with a self-build)
• Some detail is wrong (it says "NXDOMAIN" where it should be "empty NOERROR")
• OP only rules out one of the non-working solutions that are suggested at various places in the web
• I have another requirement that the OP might or might not have implied (Any solution should be non-permanent and configurable).
• The question is two years old. Given that IPv6 is enabled in more and more distros by default, more people might face this issue, and new solutions might be available. But because of its age, this question has slipped through many attention filters

I would improve the question as follows:

• Adding a number of good reasons why the question is relevant and valid.
• Rule out 2 more non-working solutions that are suggested elsewhere
• Improving title, tags and wording
• Correct some technical detail
• Adding my additional requirement
• Hinting out two rough ideas how to maybe solve it

But improving that question would have some major disadvantages:

• I would effectively re-write the question's text
• Adding my additional requirement (non-permanent solution) would sort of change the question itself
• It would still suffer from any downvotes it initially collected
• The currently only answer would still be upvoted and marked as "accepted", even though it rebuttal of the question would become obsolete, and the included solution is highly unsatisfactory.

Basically, the chances of getting fresh answers to an improved old question are slim.

Given these circumstances: Would it be legitimate to post a duplicate? If yes: How can i make sure to not get closed as duplicate?