There have been countless questions on this subject, but the closest one I found was about (something like) what defines a duplicate, the question or the answer? This one is slightly different.

I ran into [this question][1]. I saw the question was mentioned as a duplicate to [this one][2] in a comment, and started happily writing an answer in the supposed original.

Then I looked better into the exact words of the *description* of he question and read that the poster explicitly asked for a solution that also worked without GUI log in. Since my answer (not a duplicate as far as I can see) only works when logging in *with* GUI, I felt silly posting it there, and came to the conclusion that although the *question* seemed to be a duplicate, the *description* of the question asked for a different solution. Therefore it could not be an exact duplicate. I have seen other examples like this, where questions were happily closed as duplicate, while indeed the *title* suggested a duplicate, but the description made clear the origin of the question was different. 

My question is in general, since in this case there was still another duplicate that I overlooked, which *was* a real duplicate in my opinion.

So my question is: what defines a duplicate, a question, its description or both?


  [1]: http://askubuntu.com/questions/463238/how-to-run-command-at-startup/463280#463280
  [2]: http://askubuntu.com/questions/270049/how-to-run-a-command-at-login