TITLE: What do you do if you find a typo in an equation of a paper?
QUESTION [15 upvotes]: I was reading an interesting paper, and early in the introduction there was an equation with a typo in it.  I am absolutely sure of this, and all it was missing was a factor of $n!$.  Overall, it was an inconsequential part of the paper, and doesn't affect anything, as it was in the section which tells the reader about the background of the area.  But still, the equation as written is not correct, and could be confusing to someone reading who is trying to learn about the area.
In fact, the line right before this particular equation said "it is easily observed that..." which is a little interesting.

My question is:  Should I do anything when I see this?  Should I just ignore it, or is it polite to email the author? Or is it impolite to do so?  If you published a paper, with an inconsequential, but possibly confusing typo in the introduction, would you want someone to tell you about it or just leave it alone?

Thanks for the advice!  
Also please note, I am not referring to a spelling typo, that is definitely not important!  Its just possible that this typo could confuse someone reading the paper. (I was confused at first!)

REPLY [5 votes]: I usually email the author in case the paper is online in a place controlled by the author (i. e., his website or arXiv). The only downside of this approach (apart from authors not replying) is that some would just take down their papers (rather than correcting them) when they hear about a mistake in a part they consider substantial. This happened to me once (although fortunately the paper is still avaliable at other places in the net).
In case the paper is online but not in such a place, this is a more difficult question, but if the error is substantially confusing, I'd still mail the author. Unfortunately, this sometimes means notifying a clueless author that somebody else has put his paper online (most usually, a university workmate, for his students; I am not talking about pirate sites), and rather than correcting it tries to get it offline. I would be careful here.
In case the paper is offline or behind a paywall, I ignore it unless the mistake really destroys some results from the paper, in which case I rejoice about another little blow to copyrighted literature and the so-called peer review process.
But then again, I am mostly reading openly avaliable texts, so the first case is by far the most frequent.
I, personally, would prefer anyone telling me of any mistake, but I do not have (and do not plan to have) papers outside of open access, so there is no contradiction here.