Recently, I wondered what the algorithm for the placement of floats in TeX is. So I searched this site for "float placement algorithm" and did not find a suitable question - at least not directly. While starting to write my question, I noticed two questions that were suggested as related and came somewhat closer to what I was looking for and, finally, in one of these questions I found a link to How to influence the position of float environments like figure and table in LaTeX?.

Ok, so I found the answer that I was looking for, but I guess other people might also try the search phrase I used and maybe give up their search before arriving at the existing question and answer. So my question in such a situation is:

Should I ask an existing question in different words (and directly flag it as a duplicate) to improve the findability of the existing question?

As an alternative, I could of course also edit the existing question and add alternative formulations, but I guess that could easily flood the post and distract from the actual content of the question.

  • 3
    I would be against this type of tactic, even though you can do whatever you want. That's just my opinion.
    – Werner Mod
    Commented Dec 14, 2015 at 6:15
  • @Werner Could you explain the reason for your opinion? Note that my incentive is to improve the site, not to gain anything for myself (I guess by flagging the question directly, there won't be much reputation to be gained). Commented Dec 14, 2015 at 10:11
  • 14
    In-site search simply doesn't work. Try googling with the extra term site:tex.stackexchange.com to search only this site (2nd hit is the one you were looking for). It is indexed much better by google than its own search. So I would really recommend not doing this. Note that float placement and algorithm all are super generic keywords by themselves so they are not good google-fu either.
    – percusse
    Commented Dec 14, 2015 at 12:58
  • 1
    I agree with @percusse that the on-site search is rubbish. i exclusively use google if i want to find a specific question. But that means knowing that you can use site: and knowing the right keywords. Without the right search term, you won't find helpful results directly.
    – Johannes_B
    Commented Dec 14, 2015 at 13:46
  • 4
    I start asking a question if I want to search. The list of possible duplicates is much better than the results produced by searching the site. (Too much hassle to switch to a search engine when the 'ask' button' is right there.) But I don't actually ask the question, mind you. I just use starting-to-ask as a way of searching.
    – cfr
    Commented Dec 21, 2015 at 1:28
  • I even asked this question about search on meta.askubuntu....
    – Rmano
    Commented Dec 26, 2015 at 18:15

2 Answers 2


Not being able to find something has its root with searching on the site. However, this itself provides a catch-22-type problem: You need to know what to and how to search properly in order to find something meaningful. Sometimes a direct Google search (appending it with the site: tag, as in site:tex.stackexchange.com) is more effective than using the integrated SE search.

I would think a duplicate post is really something that should occur naturally, since it may just be that you couldn't find what you're looking for based on your search terms. Something that could have been avoided if you'd used more appropriate ninja search tools or Google.

Now, of course, the flip-side of this argument could be that if you had a problem finding what you're looking for, then most likely there are others who are/will be experiencing the same issue. Therefore, posting a duplicate seems logical and a benefit to the entire site/community.

Of course, there's nothing stopping you from doing as you wish and posting and flagging a direct duplicate. Duplicate questions don't fall under the automated delete criteria and should be safe from a purge.

  • I readily admit that the post was findable (I did find it after all) if one searches "the right way". I just wonder whether the idea of conciously posting a duplicate (and flagging it) to ease the search for a post is something deemed helpful/right here. I do not want to annoy people by duplicate posts if this is not seen as something helpful. Commented Dec 14, 2015 at 19:20
  • @cryingshadow: Some may find it helpful, and others may think that it is an attempt at getting some more reputation (as some users upvote as a "mark as read" feature). The best test is to do something and see what the community does. :-| You mention "duplicate posts" (plural), which makes me think you have a stash of these possible duplicates that you might want to post. And that might be problematic, since you're trying to think what other people might want... and in that case, you may make the search possibilities even more diluted.
    – Werner Mod
    Commented Dec 14, 2015 at 23:04
  • No, the plural was just a kind of generalization. I only came across this concrete situation described here, but I wanted to ask for general principles. I guess I will just try it with this particular question and see what happens. Actually, I think I directly declare it as community wiki to avoid any appearance of reputation hunting. Commented Dec 15, 2015 at 0:29
  • The new question can be found here: tex.stackexchange.com/questions/283114/… I just noticed that I can't directly mark a question as community wiki. So I left a comment indicating that this question should be turned into one. Commented Dec 15, 2015 at 0:45

An addition I just noticed: This behavior (asking a question and flagging it as a duplicate) is actually discouraged by the system since duplicate questions are counted as "bad questions" (see, e.g., https://tex.meta.stackexchange.com/a/4501/69445 or "Bad question" warning without any negative question). While it is disputable whether this classification is correct, this is the current situation and it does not seem to change in the near future.

You must log in to answer this question.

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